ATBECHDEDKESFRGBGRITLILUNLSEMCPTIESILTLVFIROMKCYALTRBGCZEEHUPLSK..HRIS..MTNORS..SM..................JBDM Ver 2.1.1 - 2999001/03352360CORRECTED EUROPEAN PATENT SPECIFICATIONB9B120240703W1deAnsprüche ENenClaims ENfrRevendications ENEP15904098.9201509172018030820190528jaenen2024070320242720180725201830202404102024152023110820240703202427

H02M 7/483 20070101AFI20230419BHEP

H02M 1/12 20060101ALI20230419BHEP

H02M 7/4835 20210501 LI20220906BHEP

H02M 1/12 20130101 FI20220906BHEP

deSTROMWANDLUNGSVORRICHTUNGenPOWER CONVERSION DEVICEfrDISPOSITIF DE CONVERSION DE PUISSANCEEP-A1- 2 887 524WO-A1-2011/042050WO-A1-2014/162620JP-A- 2012 044 839JP-A- 2013 507 100JP-A- 2014 018 028US-A1- 2014 226 373SAI PRASANNA P ET AL: "A review on circulating current suppression control, capacitor voltage balancing and fault analysis of modular multilevel converters", 2015 INTERNATIONAL CONFERENCE ON ELECTRICAL, ELECTRONICS, SIGNALS, COMMUNICATION AND OPTIMIZATION (EESCO), IEEE, 24 January 2015 (2015-01-24), pages 1-6, XP033200100, DOI: 10.1109/EESCO.2015.7253622 ISBN: 978-1-4799-7676-8 [retrieved on 2015-09-09]ZHANG MING ET AL: "Circulating Harmonic Current Elimination of a CPS-PWM-Based Modular Multilevel Converter With a Plug-In Repetitive Controller", IEEE TRANSACTIONS ON POWER ELECTRONICS, INSTITUTE OF ELECTRICAL AND ELECTRONICS ENGINEERS, USA, vol. 29, no. 4, 1 April 2014 (2014-04-01), pages 2083-2097, XP011529962, ISSN: 0885-8993, DOI: 10.1109/TPEL.2013.2269140 [retrieved on 2013-10-15]20180924FUJII, Toshiyukic/o MITSUBISHI ELECTRIC CORPORATION 7-3 Marunouchi 2-chome Chiyoda-kuTokyo 100-8310JPJIMICHI, Takushic/o MITSUBISHI ELECTRIC CORPORATION 7-3 Marunouchi 2-chome Chiyoda-kuTokyo 100-8310JPUDA, Ryosukec/o MITSUBISHI ELECTRIC CORPORATION 7-3 Marunouchi 2-chome Chiyoda-kuTokyo 100-8310JPMitsubishi Electric Corporation100785331M/FPO-176-PC/EP7-3, Marunouchi 2-chomeChiyoda-ku Tokyo 100-8310JPMeissner Bolte Partnerschaft mbB100060081Patentanwälte Rechtsanwälte Postfach 86 06 2481633 MünchenDEALATBEBGCHCYCZDEDKEEESFIFRGBGRHRHUIEISITLILTLULVMCMKMTNLNOPLPTRORSSESISKSMTRJP201507643120150917jaWO201704690920170323201712 TECHNICAL FIELD

This invention relates to a power conversion device which converts power between an alternating current (AC) and a direct current (DC) and is suitably used, for example, in a power conversion device of a large capacity installed in a power system.

BACKGROUND ART

In a power conversion device of a large capacity installed in a power system, an output from a converter has a high voltage or a high current and hence the power conversion device is often configured with a plurality of converters being multiplexed in series or in parallel. By multiplexing converters, not only an effect of increase in capacity of the converter but also an effect of lowering in harmonics contained in a waveform of an output voltage as a result of combination of output voltages from the converters and resultant lowering in harmonic current which flows out to the power system can be achieved.

Examples of the power conversion device including multiplexed converters include a multilevel converter in which output terminals of a plurality of converters are cascaded. A modular multilevel converter (MMC) represents one of the multilevel converters. The modular multilevel converter includes a first arm connected to a DC terminal on a positive-electrode side and a second arm connected to a DC terminal on a negative-electrode side for each phase of an alternating current and each arm is configured with a plurality of converter cells (which are also referred to as chopper cells) being cascaded. The first arm and the second arm of each phase implement a leg. Each leg is provided with at least one reactor.

In the modular multilevel converter, a circulating current which circulates through a plurality of legs without flowing to the outside may flow and the circulating current should be controlled to 0 or a prescribed value. Techniques described, for example, in Japanese Patent JP 5 189 105 B2 (PTD 1) and Japanese National Patent Publication JP 2012-531 878 A (PTD 2) have been known as the conventional techniques for control of a circulating current.

Japanese Patent JP 5 189 105 B2 (PTD 1) discloses a multilevel converter having one control unit for controlling and lowering a circulating current for each arm (a phase module branch). Each control unit is given a branch voltage target value from a current control unit. In particular, this document discloses combination by the current control unit of a circulating voltage target value with another target value of a phase module branch as an add-on, that is, in a linear manner, (in a form of a sum or a difference) in order to generate a branch voltage target value.

Japanese National Patent Publication JP 2012-531 878 A (PTD 2) discloses connection of a harmonic compensator of an active control type to a reactor (an inductor) provided in a leg of each phase in order to control a circulating current. This harmonic compensator is configured to suppress a harmonic component higher in frequency than a fundamental component contained in a circulating current.

Patent Document D1 is considered to be the closest prior art and relates to a voltage source converter comprising a group of phase legs, at least three connection terminals for connecting the phase legs to power transmission elements, a first group of cells in each phase leg and a second group of cells. The cells in the first group are only capable of providing unipolar voltage contributions to the converter and connected for only being capable of such unipolar voltage contributions, while the cells in the second group are connected to the corresponding cells of the first group and arranged to have bipolar voltage contribution capability.

Non Patent literature SAI PRASANNA P ET AL: "A review an circulating current suppression control, capacitor voltage balancing and fault analysis of modular multilevel converters", 2015 INTERNATIONAL CONFERENCE ON ELECTRICAL, ELECTRONICS, SIGNALS, COMMUNICATION AND OPTIMIZATION (EESCO),

IEEE, 24 January 2015 (2015-01-24), pages 1-6, XP033200100, D01: 10.1109/EESCO.2015.7253622, ISBN: 978-1-4799-7676-8 is considered relevant and it relates to a modular multilevel converters (MMCs) have proven to be the state-of-art technology due to their distinctive advantages such as low harmonic content, reduced losses, modularity and flexibility in converter design. Despite of several advantages, there are few operational challenges such as circulating current suppression control, capacitor voltage balancing and fault tolerant operation which need to be addressed. This paper presents a comprehensive review on basic operation of MMC along with different control strategies.

Patent document WO2011/042050 A1 discloses an modular multilevel converter, wherein each phase leg comprises cells belonging to a first cell group, and cells belonging to a second cell group. Cells of the first cell group are used for normal basic conversion activity, and cells of the second cell group are used for removing circulating AC currents.

CITATION LIST PATENT DOCUMENT

PTD 1: Japanese Patent JP 5 189 105 B2
PTD 2: Japanese National Patent Publication JP 2012-531 878 A

SUMMARY OF INVENTION TECHNICAL PROBLEM

More specifically, the power conversion device described in Japanese Patent JP 5 189 105 B2 (PTD 1) combines a voltage command value for control of an electric quantity (a voltage and a current) of an AC terminal, a voltage command value for control of an electric quantity (a voltage and a current) of a DC terminal, and a voltage command value for control of a circulating current which circulates in the power conversion device with one another. Then, the combined voltage command value is provided to all converter cells (chopper cells).

An upper limit and a lower limit of a voltage value which can be output by each converter cell are determined by a voltage value of a capacitor of each converter cell and a circuit configuration of each converter cell. Therefore, each converter cell cannot output a voltage exceeding the determined upper limit and lower limit. Therefore, for example, increase or decrease in voltage command value for control of electric quantities of an AC terminal and a DC terminal may restrict a voltage command value for control of a circulating current combined with these voltage command values. In this case, a voltage command value for suppressing a circulating current is not reflected on an output voltage from the converter cell. In contrast, under the influence by the voltage command value for control of a circulating current, a voltage command value for control of electric quantities of the AC terminal and the DC terminal is restricted, and consequently AC-DC conversion is not ideally achieved.

The power conversion device described in Japanese National Patent Publication No. 2012-531878 (PTD 2) is configured such that the harmonic compensator of the active control type connected to each reactor (inductor) suppresses a harmonic component higher in frequency than the fundamental component contained in the circulating current. The reactor, however, has such a characteristic that a current is more likely to flow as a frequency is lower (an admittance increases as a frequency is lower), and hence a DC current component and a fundamental component contained in the circulating current cannot be suppressed.

This invention was made in consideration of the problems described above and an object thereof is to provide a power conversion device capable of reliably controlling an AC electric quantity (an AC voltage and an AC current), a DC electric quantity (a DC voltage and a DC current), and a circulating current.

SOLUTION TO PROBLEM

According to the invention, the problem is solved by the subject-matter outlined in the independent claim. Advantageous further developments of the invention are set forth in the dependent claims.

ADVANTAGEOUS EFFECTS OF INVENTION

According to this invention, a plurality of chopper cells of each leg circuit operate in accordance with a result of a non-linear operation of a first voltage command value representing an AC voltage component to be output from the plurality of chopper cells, a second voltage command value representing a DC voltage component to be output from the plurality of chopper cells, and a third voltage command value representing a voltage to be output from the plurality of chopper cells in order to suppress a circulating current. Consequently, control of a circulating current and control of electric quantities of an AC terminal and a DC terminal can both be achieved without interference therebetween.

BRIEF DESCRIPTION OF DRAWINGS

FIG. 1: is a schematic configuration diagram of a power conversion device according to a first embodiment.
FIG. 2: is a circuit diagram showing one example of a converter cell constituting cell groups 6a, 6b, 6c, and 6d.
FIG. 3: is a configuration diagram of a control device 5 in FIG. 1.
FIG. 4: is a detailed configuration diagram of a circulating current control portion 5b in FIG. 3.
FIG. 5: is a waveform diagram schematically showing change over time of each signal in FIG. 4.
FIG. 6: is a diagram showing a configuration of circulating current control portion 5b in the power conversion device according to a second embodiment.
FIG. 7: is a diagram showing one example of relation between an input and an output of a non-linear transfer function application portion 5b22 in FIG. 6.
FIG. 8: is a diagram showing a configuration of circulating current control portion 5b in the power conversion device according to a third embodiment.
FIG. 9: is a diagram showing a configuration of circulating current control portion 5b in the power conversion device according to a fourth embodiment.
FIG. 10: is a diagram showing one example of relation between inputs x and y and an output z of a non-linear mathematical function application portion in FIG. 9.
FIG. 11: is a diagram showing a configuration of circulating current control portion 5b in the power conversion device according to a fifth embodiment.
FIG. 12: is a diagram schematically showing a waveform showing change over time of a voltage command value Vpref2 output from circulating current control portion 5b.
FIG. 13: is a schematic configuration diagram of a power conversion device according to a sixth embodiment.
FIG. 14: is a circuit diagram showing a detailed configuration of each cell 20 provided in cell groups 6c and 6d for control of a circulating current.
FIG. 15: is a configuration diagram of control device 5 in FIG. 13.
FIG. 16: is a configuration diagram of control device 5 included in the power conversion device according to a seventh embodiment.

DESCRIPTION OF EMBODIMENTS

Each embodiment will be described below in detail with reference to the drawings. The same or corresponding elements have the same reference characters allotted and description thereof will not be repeated.

First Embodiment Schematic Configuration of Power Conversion Device

FIG. 1 is a schematic configuration diagram of a power conversion device according to a first embodiment. Referring to FIG. 1, the power conversion device includes leg circuits 8a, 8b, and 8c which are main circuits (which are denoted as a leg circuit 8 when they are collectively referred to or an unspecified one is referred to) and a control device 5 which controls these leg circuits 8.

Leg circuit 8 is provided for each of a plurality of phases implementing an alternating current and converts power between an alternating current and a direct current. FIG. 1 shows an example of a three-phase alternating current, and three leg circuits 8a, 8b, and 8c corresponding to a u phase, a v phase, and a w phase, respectively, are provided.

AC terminals Nu, Nv, and Nw provided in respective leg circuits 8a, 8b, and 8c are connected to AC circuit 2 with an interconnected transformer 3 being interposed.
AC circuit 2 is, for example, an AC power system including an AC power supply. For facilitating illustration, FIG. 1 does not show connection of AC terminals Nv and Nw to interconnected transformer 3. DC terminals Np and Nn (a positive-side DC terminal Np and a negative-side DC terminal Nn) provided in common to leg circuits 8 are connected to DC circuit 4. DC circuit 4 is, for example, a DC power system and includes a DC power grid and another power conversion device which provides a DC output.

Instead of interconnected transformer 3 in FIG. 1, AC terminals Nu, Nv, and Nw may be connected to AC circuit 2 with an interconnected reactor being interposed. Instead of AC terminals Nu, Nv, and Nw, a primary winding may be provided in each of leg circuits 8a, 8b, and 8c, and leg circuits 8a, 8b, and 8c may be connected in an AC manner to interconnected transformer 3 or an interconnected reactor with a secondary winding magnetically coupled to the primary winding being interposed. In this case, the primary winding may be implemented by reactors 7a and 7b. Each of leg circuits 8a, 8b, and 8c is electrically (in a DC or AC manner) connected to AC circuit 2 with a connection portion provided in each of leg circuits 8a, 8b, and 8c being interposed, such as AC terminals Nu, Nv, and Nw or the above-described primary winding.

Leg circuit 8a is divided into a positive-side arm (which is also referred to as an upper arm or a primary arm) 13 from positive-side DC terminal Np to AC input terminal Nu and a negative-side arm (which is also referred to as a lower arm or a secondary arm) 14 from negative-side DC terminal Nn to AC input terminal Nu. A point of connection Nu between positive-side arm 13 and negative-side arm 14 is connected to transformer 3. Positive-side DC terminal Np and negative-side DC terminal Nn are connected to DC circuit 4. Since leg circuits 8b and 8c are also similarly configured, leg circuit 8a will be described below as a representative.

Positive-side arm 13 includes a cell group 6a in which a plurality of converter cells (chopper cells) 1 are cascaded, a cell group 6c in which a plurality of converter cells 1 are cascaded, and reactor 7a. Cell groups 6a and 6c and reactor 7a are connected in series to one another. For the sake of brevity, a converter cell (chopper cell) may be referred to as a cell below. Though FIG. 1 shows only a single cell 1 in cell group 6c for facilitating illustration, a plurality of cells 1 are actually cascaded.

Similarly, negative-side arm 14 includes a cell group 6b in which a plurality of cells 1 are cascaded, a cell group 6d in which a plurality of cells 1 are cascaded, and reactor 7b. Cell groups 6b and 6d and reactor 7b are connected in series to one another. Though FIG. 1 shows only a single cell 1 in cell group 6d for facilitating illustration, a plurality of cells 1 are actually cascaded.

Reactor 7a may be inserted in any position in positive-side arm 13 of leg circuit 8a, and reactor 7b may be inserted in any position in negative-side arm 14 of leg circuit 8a.
A plurality of reactors 7a and a plurality of reactors 7b may be provided. The reactors may be different in inductance value. Only reactor 7a of positive-side arm 13 or only reactor 7b of negative-side arm 14 may be provided.

Cell groups 6a and 6c provided in positive-side arm 13 are referred to as a positive-side cell group and cell groups 6b and 6d provided in negative-side arm 14 are referred to as a negative-side cell group. As will be described in detail below, positive-side cell group 6a and negative-side cell group 6b are not used for control of a circulating current but are used only for control of an AC electric quantity and a DC electric quantity. Positive-side cell group 6c and negative-side cell group 6d are used for control of a circulating current.
Control of a circulating current is characterized in that only at least one cell constituting each leg circuit 8 is used therefor.

The power conversion device in FIG. 1 further includes an AC voltage detector 10, DC voltage detectors 11a and 11b, and arm current detectors 9a and 9b provided in each leg circuit 8 as detectors which measure electric quantities (a current and a voltage) used for control. Signals detected by these detectors are input to control device 5.

Specifically, AC voltage detector 10 detects a U-phase voltage value Vacu, a V-phase voltage value Vacv, and a W-phase voltage value Vacw of AC circuit 2.
DC voltage detector 11a detects a voltage of positive-side DC terminal Np connected to DC circuit 4. DC voltage detector 11b detects a voltage of negative-side DC terminal Nn connected to DC circuit 4. Arm current detectors 9a and 9b provided in leg circuit 8a for the U phase detect an arm current Ipu which flows in positive-side arm 13 and an arm current Inu which flows in negative-side arm 14, respectively. Similarly, arm current detectors 9a and 9b provided in leg circuit 8b for the V phase detect a positive-side arm current Ipv and a negative-side arm current Inv, respectively. Arm current detectors 9a and 9b provided in leg circuit 8c for the W phase detect a positive-side arm current Ipw and a negative-side arm current Inw, respectively. Arm currents Ipu, Inu, Ipv, Inv, Ipw, and Inw which flow from positive-side DC terminal Np toward negative-side DC terminal Nn are defined as positive currents.

Configuration Example of Converter Cell

FIG. 2 is a circuit diagram showing one example of a converter cell constituting cell groups 6a, 6b, 6c, and 6d. Converter cell 1 shown in FIG. 2(a) adopts a half bridge configuration and includes semiconductor switching elements 1a and 1b (which may hereinafter simply be referred to as a switching element) connected in series to each other, diodes 1c and 1d, and a DC capacitor 1e. Diodes 1c and 1d are connected in anti-parallel (in parallel and in a direction of a reverse bias) to switching elements 1a and 1b, respectively. DC capacitor 1e is connected in parallel to a series connection circuit of switching elements 1a and 1b and smoothes a DC voltage. A connection node between switching elements 1a and 1b is connected to a positive-side input and output terminal 1p and a connection node between switching element 1b and DC capacitor 1e is connected to a negative-side input and output terminal 1n.

In the configuration in FIG. 2(a), switching elements 1a and 1b are controlled such that one is turned on and the other is turned off. When switching element 1a is turned on and switching element 1b is turned off, a voltage across opposing ends of DC capacitor 1e is applied across input and output terminals 1p and 1n (a positive-side voltage being applied to input and output terminal 1p and a negative-side voltage being applied to input and output terminal 1n). In contrast, when switching element 1a is turned off and switching element 1b is turned on, 0 V is applied across input and output terminals 1p and 1n. Converter cell 1 shown in FIG. 2(a) can output a zero voltage or a positive voltage (dependent on a voltage of DC capacitor 1e) by alternately turning on switching elements 1a and 1b. Diodes 1c and 1d are provided for protection when voltages in reverse directions are applied to switching elements 1a and 1b.

Converter cell 1 shown in FIG. 2(b) adopts a full bridge configuration and it is different from converter cell 1 in FIG. 2(a) in further including switching elements 1f and 1g connected in series and diodes 1h and 1i connected in anti-parallel to switching elements 1f and 1g. Switching elements 1f and 1g as a whole are connected in parallel to the series connection circuit of switching elements 1a and 1b and connected in parallel to DC capacitor 1e. Input and output terminal 1p is connected to the connection node between switching elements 1a and 1b and input and output terminal 1n is connected to a connection node between switching elements 1f and 1g.

Converter cell 1 shown in FIG. 2(b) is controlled such that switching element 1g is normally turned on, switching element 1f is normally turned off, and switching elements 1a and 1b are alternately turned on during a normal operation (that is, a zero voltage or a positive voltage is output across input and output terminals 1p and 1n). Converter cell 1 shown in FIG. 2(b), however, can also output a zero voltage or a negative voltage by turning off switching element 1g, turning on switching element 1f, and alternately turning on switching elements 1a and 1b.

Converter cell 1 shown in FIG. 2(c) is configured with switching element 1f having been removed from converter cell 1 of the full bridge configuration as shown in FIG. 2(b) and it is otherwise the same as in FIG. 2(b). Converter cell 1 in FIG. 2(c) is controlled such that switching element 1g is normally turned on and switching elements 1a and 1b are alternately turned on during the normal operation (that is, a zero voltage or a positive voltage is output across input and output terminals 1p and 1n). Converter cell 1 shown in FIG. 2(c) can output a negative voltage when switching elements 1a and 1g are turned off and switching element 1b is turned on so that a current flows from input and output terminal 1n toward input and output terminal 1p.

A self-turn-off switching element capable of control of both of an on operation and an off operation is employed for each of switching elements 1a, 1b, 1f, and 1g. For example, an insulated gate bipolar transistor (IGBT) or a gate commutated turn-off thyristor (GCT) is employed as switching elements 1a, 1b, 1f, and 1g.

Configuration and General Operation of Control Device

FIG. 3 is a configuration diagram of control device 5 in FIG. 1. Control device 5 shown in FIG. 3 may be configured with a dedicated circuit or configured in part or in its entirety with a field programmable gate array (FPGA) and/or a microprocessor. A configuration of control device 5 and a general operation of each element will be described below with reference to FIGS. 1 and 3.

Control device 5 includes a voltage command value generation portion 5z and gate control portions 5k, 5m, 5n, and 5o. Gate control portion 5k supplies gate signals Gpu, Gpv, and Gpw to each switching element constituting positive-side cell group 6a of leg circuits 8a, 8b, and 8c, respectively. Gate control portion 5m supplies gate signals Gnu, Gnv, and Gnw to each switching element constituting negative-side cell group 6b of leg circuits 8a, 8b, and 8c, respectively. Gate control portion 5n supplies gate signals Gp2u, Gp2v, and Gp2w to each switching element constituting positive-side cell group 6c for control of a circulating current of leg circuits 8a, 8b, and 8c, respectively. Gate control portion 5o supplies gate signals Gn2u, Gn2v, and Gn2w to each switching element constituting negative-side cell group 6d for control of a circulating current of leg circuit 8a, 8b, and 8c, respectively.

Voltage command value generation portion 5z supplies voltage command values Vpref, Vnref, Vpref2, and Vnref2 to gate control portions 5k, 5m, 5n, and 5o, respectively. Voltage command values Vpref2 and Vnref2 supplied to respective gate control portions 5n and 5o for control of a circulating current are based on a detection value of a circulating current Icc. Voltage command values Vpref and Vnref supplied to other gate control portions 5k and 5m are not based on a detection value of circulating current Icc.

More specifically, voltage command value generation portion 5z includes a current operation portion 5a, a circulating current control portion 5b, an AC control portion 5c, a DC control portion 5d, command value combination portions 5e and 5f, and gain circuits 5g and 5h.

Current operation portion 5a takes in positive-side arm currents Ipu, Ipv, and Ipw detected by current detector 9a provided in positive-side arm 13 of leg circuit 8 of each phase and negative-side arm currents Inu, Inv, and Inw detected by current detector 9b provided in negative-side arm 14 of leg circuit 8 of each phase. Current operation portion 5a operates AC current values Iacu, Iacv, and Iacw, a DC current value Idc, and circulating current values Iccu, Iccv, and Iccw from the arm current which has been taken in. Current operation portion 5a outputs calculated AC current values Iacu, Iacv, and Iacw to AC control portion 5c, outputs calculated DC current value IDC to DC control portion 5d, and outputs calculated circulating current values Iccu, Iccv, and Iccw to circulating current control portion 5b.

U-phase AC current Iacu, V-phase AC current Iacv, and W-phase AC current Iacw (which are denoted as an AC current IAC when they are collectively referred to) which flow from AC terminals Nu, Nv, and Nw of each leg circuit 8 toward transformer 3 are defined as positive. DC current IDC which flows from DC circuit 4 toward positive-side DC terminal Np and from negative-side DC terminal Nn toward DC circuit 4 is defined as positive. Circulating currents Iccu, Iccv, and Iccw which flow through leg circuits 8a, 8b, and 8c, respectively (which are denoted as circulating current Icc when they are collectively referred to) from positive-side DC terminal Np toward negative-side DC terminal Nn are defined as positive.

AC voltage values Vacu, Vacv, and Vacw of the U phase, the V phase and the W phase (which are denoted as an AC voltage value VAC when they are collectively referred to) detected by AC voltage detector 10 are further input to AC control portion 5c.
AC control portion 5c generates AC voltage command values Vacrefu, Vacrefv, and Vacrefw of the U phase, the V phase, and the W phase (which are denoted as an AC voltage command value Vacref when they are collectively referred to) based on input AC current value IAC and AC voltage value Vac. AC voltage command value Vacref represents an AC voltage component to be output from cell groups 6a and 6b.

DC voltage values Vdcp and Vdcn detected by DC voltage detectors 11a and 11b are further input to DC control portion 5d. DC control portion 5d generates a DC voltage command value Vdcref based on input DC voltage values Vdcp and Vdcn and DC current value Idc. DC voltage command value Vdcref represents a DC voltage component to be output from cell groups 6a and 6b.

Command value combination portion 5e generates a voltage command value Vprefu for U-phase positive-side cell group 6a by combining U-phase AC voltage command value Vacrefu and DC voltage command value Vdcref with each other. Similarly, command value combination portion 5e generates a voltage command value Vprefv for V-phase positive-side cell group 6a by combining V-phase AC voltage command value Vacrefv and DC voltage command value Vdcref with each other. Command value combination portion 5e further generates a voltage command value Vprefw for W-phase positive-side cell group 6a by combining W-phase AC voltage command value Vacrefw and DC voltage command value Vdcref with each other. Generated voltage command values Vprefu, Vprefv, and Vprefw (which are denoted as voltage command value Vpref when they are collectively referred to or an unspecified voltage command value is referred to) is input to gate control portion 5k.

Command value combination portion 5f generates a voltage command value Vnrefu for U-phase negative-side cell group 6b by combining U-phase AC voltage command value Vacrefu and DC voltage command value Vdcref with each other. Similarly, command value combination portion 5f generates a voltage command value Vnrefv for V-phase negative-side cell group 6b by combining V-phase AC voltage command value Vacrefv and DC voltage command value Vdcref with each other. Command value combination portion 5f further generates a voltage command value Vnrefw for W-phase negative-side cell group 6b by combining W-phase AC voltage command value Vacrefw and DC voltage command value Vdcref with each other. Generated voltage command values Vnrefu, Vnrefv, and Vnrefw (which are denoted as voltage command value Vnref when they are collectively referred to or an unspecified voltage command value is referred to) are input to gate control portion 5m.

Gain circuit 5g outputs a value calculated by multiplying voltage command value Vpref for positive-side cell group 6a of each phase by gain K to circulating current control portion 5b as a positive-side voltage correction value Vpcorr for control of a circulating current (Vpcorru for the U phase, Vpcorrv for the V phase, and Vpcorrw for the W phase). Gain circuit 5h outputs a value calculated by multiplying voltage command value Vnref for a negative-side cell group 6b of each phase by gain K to circulating current control portion 5b as a negative-side voltage correction value Vncorr for control of a circulating current (Vncorru for the U phase, Vncorrv for the V phase, and Vncorrw for the W phase).

Circulating current control portion 5b generates voltage command value Vpref2 for positive-side cell group 6c (Vpref2,u for the U phase, Vpref2v for the V phase, and Vpref2w for the W phase) based on circulating current value Ice from current operation portion 5a and positive-side voltage correction value Vpcorr for each phase. Circulating current control portion 5b further generates voltage command value Vnref2 for negative-side cell group 6d (Vnref2,u for the U phase, Vnref2v for the V phase, and Vnref2w for the W phase) based on circulating current value Ice from current operation portion 5a and negative-side voltage correction value Vncorr for each phase. Generated voltage command value Vpref2 is supplied to gate control portion 5n and generated voltage command value Vnref2 is supplied to gate control portion 5o.

Overview of Configuration and Operation of Circulating Current Control Portion 5b

FIG. 4 is a detailed configuration diagram of circulating current control portion 5b in FIG. 3. Referring to FIG. 4, circulating current control portion 5b includes a subtractor 5b9, a compensator 5b18, an adder 5b19, and multipliers 5b20 and 5b21.

Subtractor 5b9 calculates a difference between a circulating current command value Iccref (Iccrefu for the U phase, Iccrefv for the V phase, and Iccrefw for the W phase) and circulating current value Ice (Iccu of the U phase, Iccv of the V phase, and Iccw of the W phase) by subtracting the circulating current value from the circulating current command value. Compensator 5b18 generates voltage command value Vccref for control of a circulating current (Vccrefu for the U phase, Vccrefv for the V phase, and Vccrefw for the W phase) by amplifying the calculated difference. Though a zero current is normally provided as a circulating current command value, a non-zero value may also be provided when imbalance occurs in a power system.

Adder 5b19 adds a fixed value to the amplified difference (that is, voltage command value Vccref). Though the fixed value is set to 1.0 in FIG. 4, the fixed value is not limited to 1.0. Multiplier 5b20 generates voltage command value Vpref2 for positive-side cell group 6c by multiplying an output value from adder 5b19 (that is, Vccref + 1) by positive-side voltage correction value Vpcorr for each phase. Multiplier 5b21 generates voltage command value Vnref2 for negative-side cell group 6d by multiplying an output value from adder 5b19 (Vccref + 1) by negative-side voltage correction value Vncorr for each phase.

Circulating current control portion 5b thus generates voltage command values Vpref2 and Vnref2 for controlling cell groups 6c and 6d for control of a circulating current only through a non-linear operation of voltage command value Vccref based on circulating current Icc and voltage command values Vpref and Vnref which are based on AC current IAC and DC current IDC but are not based on circulating current Icc (or AC voltage command value Vacref and DC voltage command value Vdcref which serve as the basis) (that is, without through linear combination).

Detailed Operation of Control Device 5

A detailed operation of control device 5 will now be described.

(Operation of Current Operation Portion 5a)

Referring to FIG. 1, a point of connection between positive-side arm 13 and negative-side arm 14a of leg circuit 8a of the U phase is AC terminal Nu, which is connected to transformer 3. Therefore, AC current Iacu which flows from AC terminal Nu toward transformer 3 is equal to a current value calculated by subtracting a value of current Ipn which flows through negative-side arm 14 measured with current detector 9b from a value of current Inu which flows through positive-side arm 13 measured with current detector 9a, that is, $Iacu = Ipu - Inu$

An average value of current Ipu which flows through positive-side arm 13 and current Inu which flows through negative-side arm 14 is defined as a common current Icomu which flows through both of arms 13 and 14. Common current Icomu is a leg current which flows through a DC terminal of leg circuit 8a. Leg current Icomu can be operated as $Icomu = (Ipu + Inu) / 2$

Similarly for the V phase and the W phase, V-phase AC current Iacv and a V-phase leg current Icomv can be calculated with V-phase positive-side arm current Ipv and V-phase negative-side arm current Inv, and W-phase AC current Iacw and a W-phase leg current Icomw can be calculated with W-phase positive-side arm current Ipw and W-phase negative-side arm current Inw. Specifically, they are expressed in expressions below. $Iacv = Ipv - Inv$ $Icomv = (Ipv + Inv) / 2$ $Iacw = Ipw - Inw$ $Icomw = (Ipw + Inw) / 2$

DC terminals on the positive side of leg circuits 8a, 8b, and 8c of respective phases are connected in common as positive-side DC terminal Np, and DC terminals on the negative side are connected in common as negative-side DC terminal Nn. According to this configuration, a current value calculated by adding leg currents Icomu, Icomv, and Icomw of respective phases results in DC current IDC which flows in from the positive-side terminal of DC circuit 4 and returns to DC circuit 4 through the negative-side terminal. Therefore, DC current IDC can be operated as $IDC = Icomu + Icomv + Icomw$

Uniform allocation of a DC current component contained in the leg current among the phases is appropriate because a current capacity of the cell can be uniform. Taking into account this fact, a difference between a leg current and 1/3 of a DC current value can be operated as a value of a circulating current which does not flow in DC circuit 4 but flows through the legs of the respective phases. Specifically, circulating currents Iccu, Iccv, and Iccw of the U phase, the V phase, and the W phase can be operated as $Iccu = Icomu - Idc / 3$ $Iccv = Icomv - Idc / 3$ $Iccw = Icomw - Idc / 3$

Current operation portion 5a in FIG. 3 operates AC current values Iacu, Iacv, and Iacw, DC current value Idc, and circulating current values Iccu, Iccv, and Iccw from arm current values Ipu, Inu, Ipv, Inv, Ipw, and Inw detected by current detectors 9a and 9b in accordance with the expressions (1), (3), (5), and (7) to (10). Current operation portion 5a outputs calculated AC current values Iacu, Iacv, and Iacw, DC current value Idc, and circulating current values Iccu, Iccv, and Iccw to AC control portion 5c, DC control portion 5d, and circulating current control portion 5b, respectively.

AC control portion 5c outputs an AC voltage to be output from cell 1 of the power conversion device as AC voltage command values Vacrefu, Vacrefv, and Vacrefw, based on AC voltage values Vacu, Vacv, and Vacw detected by AC voltage detector 10 and AC current values Iacu, Iacv, and Iacw output from current operation portion 5a. In AC control portion 5c, for example, AC current control which carries out feedback control such that an AC current value matches with an AC current command value or AC voltage control which provides feedback such that an AC voltage value matches with an AC voltage command value is configured in accordance with a function required in the power conversion device. Alternatively, power control which calculates power based on an AC current value and an AC voltage value and a power value attains to a desired value is implemented. One type or a plurality of types of such control as combined is/are implemented and operated. Since AC current control controls a current output to AC circuit 2 with transformer 3 being interposed, a voltage component for controlling the current is a component which has been known as a normal-phase component and a reverse-phase component of a multi-phase AC voltage or a normal mode component. AC voltage control also similarly outputs a normal-phase component and a reverse-phase component. When an AC multi-phase voltage is output, output of a voltage component in common among the three phases which is known as a zero-phase component or a common mode component in addition to these normal- and reverse-phase components is also required. For example, it has been known that an AC component of a fundamental wave which can be output from a converter cell can be increased by approximately 15% by superimposing a third-order harmonic of the fundamental wave on a zero-phase component. In the power conversion device configured in FIG. 1, as will be shown later as outputting a certain zero-phase component, AC voltage components output from cell groups 6a and 6b are opposite in polarity to each other. Therefore, a DC component of a voltage from cell group 6a of each phase configured at a positive electrode and a DC component of a voltage output from cell group 6b of each phase configured at a negative electrode are superimposed in a direction reverse in terms of positive and negative. Consequently, a difference between DC power output from the positive-side cell group and DC power from the negative-side cell group is produced, and therefore energy in DC capacitor 1e configured in converter cell 1 can be exchanged between the positive side and the negative side, which also contributes to control for balancing between voltage values of the DC capacitors of cell group 6a and cell group 6b.

Operation of DC Control Portion 5d

DC control portion 5d operates voltage value VDC across DC terminals based on a differential voltage between DC voltage values Vdcp and Vdcn detected by DC voltage detectors 11a and 11b. Voltage value VDC across the DC terminals is given as $VDC = Vdcp - Vdcn$ DC control portion 5d generates as DC voltage command value Vdcref, a DC voltage to be output from cell 1 based on calculated voltage value VDC across the DC terminals and DC current value IDC output from current operation portion 5a, and outputs the DC voltage command value.

DC control portion 5d is implemented, for example, by any one of a DC current controller which controls a DC current value, a DC voltage controller which controls a DC voltage, and a DC power controller which controls DC power, or a plurality of them as being combined for operation, similarly to AC control portion 5c. A DC voltage component output from positive-side cell group 6a and a DC voltage component output from negative-side cell group 6b in accordance with DC voltage command value Vdcref output from the DC voltage controller, the DC current controller, and the DC power controller are identical in polarity to each other as will be described later. Since cell groups 6a and 6b are connected in series, output voltages from cell groups 6a and 6b are combined and the combined voltage is defined as a voltage component generated across the positive-side DC terminal and the negative-side DC terminal of leg circuit 8. Since DC voltage command value Vdcref is provided to gate control portions 5k and 5m as a component common among the phases in the configuration of control device 5 shown in FIG. 3, the voltage components output from cell groups 6a and 6b in accordance with DC voltage command value Vdcref result in a DC voltage component output to DC circuit 4.

Unlike the above, DC control portion 5d can also be configured to provide DC voltage command values Vdcref different in magnitude among the phases. In that case, a DC voltage command value is provided such that a circulating current which circulates among the phases flows based on a potential difference produced in reactors 7a and 7b. When a circulating current flows in a DC manner, a difference is produced in DC power generated by leg circuits 8a, 8b, and 8c and consequently a difference in energy stored in DC capacitor 1e constituting cell groups 6a and 6b is also produced among the phases. This operation is applied to balance control for balancing among the phases in connection with a DC voltage of DC capacitor 1e.

Operation of Command Value Combination Portions 5e and 5f

Command value combination portion 5e operates a voltage to be output from positive-side cell group 6a as voltage command value Vpref (Vprefu, Vprefv, and Vprefw). Command value combination portion 5f operates a voltage to be output from negative-side cell group 6b as voltage command value Vnref (Vnrefu, Vnrefv, and Vnrefw). Voltage command values Vpref and Vnref are obtained by combining DC voltage command value Vdcref and AC voltage command value Vacref with each other for each phase.

Specifically, positive-side cell group 6a and negative-side cell group 6b are connected in series between DC terminals Np and Nn connected to DC circuit 4. Therefore, in calculating each of voltage command value Vpref of positive-side cell group 6a and voltage command value Vnref of negative-side cell group 6b, 1/2 of DC voltage command value Vdcref is combined by addition.

Since AC terminals Nu, Nv, and Nw are located at points of connection between positive-side arm 13 and negative-side arm 14, AC voltage command value Vacref is combined by subtraction in calculating voltage command value Vpref of positive-side cell group 6a, and AC voltage command value Vacref is combined by addition in calculating voltage command value Vnref of negative-side cell group 6b. For example, in leg circuit 8a in FIG. 1, when positive-side cell group 6a outputs an AC voltage relatively small in value and negative-side cell group 6b outputs an AC voltage relatively great in value, a potential of AC terminal Nu is closer to a potential of positive-side DC terminal Np and a high voltage is output to AC terminal Nu. Negative-side cell group 6b outputs an AC voltage identical in polarity to an AC voltage to be output from AC terminal Nu, and positive-side cell group 6a outputs an AC voltage opposite in polarity to an AC voltage to be output from AC terminal Nu.

In the power conversion device in the first embodiment, command value combination portions 5e and 5f combine normal- and reverse-phase components and a zero-phase component contained in AC voltage command value Vacref with DC voltage command value Vdcref through the operation above, however, they do not combine a voltage component with which energy is balanced among the phases by feeding a circulating current nor a voltage component with which a circulating current is controlled.

Operation of Gate Control Portions 5k and 5m

Gate control portion 5k provides corresponding gate signals Gpu, Gpv, and Gpw to the switching element of cell 1 constituting positive-side cell group 6a of each phase based on voltage command values Vprefu, Vprefv, and Vprefw of the U phase, the V phase, and the W phase combined by command value combination portion 5e. Gate control portion 5m provides corresponding gate signals Gnu, Gnv, and Gnw to the switching element of cell 1 constituting negative-side cell group 6b of each phase based on voltage command values Vnrefu, Vnrefv, and Vnrefw of the U phase, the V phase, and the W phase combined by command value combination portion 5f.

As described already, in half bridge cell 1 shown in FIG. 2(a), switching element 1a is turned on and switching element 1b is turned off when a voltage of DC capacitor 1e is output. When a zero voltage is output, in contrast, switching element 1a is turned off and switching element 1b is turned on. A pulse width modulation (PWM) scheme has thus been known as a scheme for controlling a converter which can output a binary voltage level.

Under the pulse width modulation scheme, a pulse width of a gate signal supplied to a switching element is controlled such that a DC component of a desired voltage or an AC component of a fundamental wave can be output in a time average manner. By making pulses from a plurality of converters different in timing, a voltage less in harmonic component can be supplied as a combined voltage. For example, a method of determining timing of switching at a point of intersection between signals based on comparison of a triangular wave or a saw tooth wave of a fixed frequency with a voltage command value has been known.

Operation of Circulating Current Control Portion 5b

Circulating current values Iccu, Iccv, and Iccw of the U phase, the V phase, and the W phase operated by current operation portion 5a are sent to circulating current control portion 5b. Circulating current control portion 5b subjects the circulating current value to feedback control such that the circulating current value matches with a circulating current command value. Therefore, circulating current control portion 5b is provided with compensator 5b18 which amplifies a difference between a circulating current command value and a circulating current value. Though a zero current is normally provided as a circulating current command value, a non-zero value may also be provided when imbalance occurs in a power system. Circulating current control portion 5b outputs a voltage component to be output by cell groups 6c and 6d for control of a circulating current as voltage command value Vccref (Vccrefu for the U phase, Vccrefv for the V phase, and Vccrefw for the W phase).

The circulating current flows through legs of different phases. Cell groups 6a and 6b and reactors 7a and 7b are present in a path for the circulating current and a circulating current is produced by application of a potential difference produced by switching of cell groups 6a and 6b to reactors 7a and 7b. Therefore, a circulating current is suppressed by application of a voltage opposite in polarity to the reactors by cell groups 6c and 6d provided in the same path.

For example, when circulating current Iccu flows from the positive-side
DC terminal toward the negative-side DC terminal of leg circuit 8a and when a positive voltage is output from each of cell groups 6c and 6d of leg circuit 8a, a voltage in a direction of lowering in circulating current is applied to reactors 7a and 7b. When a current flows in a direction reverse to the above, a circulating current can be attenuated by applying a voltage from cell groups 6c and 6d also in the reverse direction. Therefore, as shown in FIG. 4, circulating current control portion 5b carries out feedback control by using compensator 5b18 which amplifies a difference between the circulating current command value and the circulating current value.

Since circulating currents Iccu, Iccv, and Iccw which flow through respective leg circuits 8 have positive and negative polarities, voltage command value Vccref output from compensator 5b18 also has positive and negative polarities. In contrast, when converter cell 1 configured as a half bridge shown in FIG. 2(a) is employed in cell groups 6c and 6d for control of a circulating current, converter cell 1 can output only a zero voltage or a positive voltage (a voltage value of the capacitor), and hence such an unfavorable condition should be avoided. In this case, if a DC bias signal is simply added to voltage command value Vccref, DC capacitor 1e of converter cell 1 will disadvantageously be kept charged because a DC current Id flows between leg circuits 8a, 8b, and 8c and DC circuit 4 in the power conversion device configured in FIG. 1. The power conversion device in the first embodiment thus adopts circulating current control portion 5b configured as shown in FIG. 4. An operation of circulating current control unit 5b is described in further detail below.

In general, in a circuit scheme of the power conversion device shown in FIGS. 1 and 2, each converter cell 1 has been known to be controlled such that energy which flows in and out of DC capacitor 1e is substantially zero. In order to achieve this, a command value for AC control and a command value for DC control are provided to each converter cell 1 such that AC power which flows in and DC power which flows out match with each other or AC power which flows out and DC power which flows in match with each other, which means that, when each converter cell 1 constituting positive-side cell group 6a is controlled with voltage command value Vpref, active power which flows in or out of each converter cell 1 is substantially zero under a current condition at that time (magnitude and a phase of an AC current, a DC current, and a circulating current).

Therefore, even though a signal in proportion to voltage command value Vpref for positive-side cell group 6a is provided to positive-side cell group 6c for control of a circulating current, positive-side cell group 6a and positive-side cell group 6c are equal to each other in current condition and therefore active power which flows in or out of each converter cell 1 constituting positive-side cell group 6c can be substantially 0. Since voltage command value Vpref2 for control of a circulating current provided to positive-side cell group 6c serves for control of a voltage to be applied at reactors 7a and 7b, power which flows in or out of each cell 1 constituting positive-side cell group 6c based on voltage command value Vpref2 is mainly composed of reactive power. This is also applicable to negative-side cell group 6d for control of a circulating current. Essentially, cell groups 6c and 6d for control of a circulating current have to output substantially no active power.

From the point of view above, multiplier 5b20 in FIG. 4 multiplies positive-side voltage correction value Vpcorr representing a value in proportion to voltage command value Vpref for positive-side cell group 6a by an output value from adder 5b19 (a result of addition of 1 to voltage command value Vccref output from compensator 5b18). The output value from multiplier 5b20 is used as voltage command value Vpref2 for positive-side cell group 6c. In this case, an output value from each converter cell 1 constituting positive-side cell group 6c is increased or decreased from a condition of active power of zero in accordance with the output value from compensator 5b18 (voltage command value Vccref). When the output value from compensator 5b18 is positive, the output value from adder 5b19 is greater than 1.0, whereas when the output value from compensator 5b18 is negative, the output value from adder 5b19 is smaller than 1.0. Therefore, positive-side voltage correction value Vpcorr can be modified in accordance with the output value from compensator 5b18. A circulating current can be controlled by setting modified positive-side voltage correction value Vpcorr as voltage command value Vpref2 to be output from positive-side cell group 6c. The above is also applicable to voltage control of each cell 1 in negative-side cell group 6d.

FIG. 5 is a waveform diagram schematically showing change over time of each signal in FIG. 4. The waveform diagram in FIG. 5 representatively shows a waveform of one phase of a three-phase alternating current.

Referring to FIG. 5, positive-side voltage correction value Vpcorr (a chain dotted line) and negative-side voltage correction value Vncorr (a dashed line) are in proportion to voltage command values Vpref and Vnref for cell groups 6a and 6b generated by command value combination portions 5e and 5f in FIG. 3, respectively. Command value combination portion 5f combines AC voltage command value Vacref output from AC control portion 5c as it is with DC voltage command value Vdcref output from DC control portion 5d, whereas command value combination portion 5e combines AC voltage command value Vacref with DC voltage command value Vdcref with a polarity thereof being inverted. Therefore, AC components of voltage command values Vpref and Vnref for cell groups 6a and 6b are opposite in polarity to each other, and AC components of voltage correction values Vpcorr and Vncorr for control of a circulating current are also opposite in polarity to each other.

Voltage command values Vpref2 and Vnref2 for cell groups 6c and 6d are obtained by multiplying an output value from adder 5b19 (that is, Vccref + 1) by voltage correction values Vpcorr and Vncorr, respectively. Therefore, sensitivity of the output value from compensator 5b18 (that is, voltage command value Vccref for control of a circulating current) to voltage command values Vpref2 and Vnref2 is dependent on magnitude of voltage correction values Vpcorr and Vncorr. Consequently, as shown in FIG. 5, sensitivity of voltage command value Vccref to voltage command value Vpref2 for positive-side cell group 6c and sensitivity of voltage command value Vccref to voltage command value Vnref2 for negative-side cell group 6d repeat increase and decrease for each half cycle of an AC period. Specifically, in a first half cycle of the alternating current in FIG. 5 (from time t1 to time t2), a circulating current is controlled mainly in negative-side cell group 6d so as to match with a command value thereof, and a voltage of DC capacitor 1e is controlled to be constant in value in positive-side cell group 6c. In a remaining half cycle in FIG. 5 (from time t2 to time t3), a circulating current is controlled mainly in positive-side cell group 6c so as to match with a command value thereof and a voltage of DC capacitor 1e is controlled to be constant in value in negative-side cell group 6d. As an operation to control a circulating current is thus alternately performed in positive-side cell group 6c and negative-side cell group 6d for each half cycle of the alternating current, control of the circulating current and maintaining of a voltage of the DC capacitor can both be achieved.

Proportional gain K is set to any such value as not saturating an output voltage from converter cell 1 when voltage command value Vccref for control of a circulating current is provided. When converter cell 1 configured as a full bridge shown in FIG. 2(b) constitutes each cell 1 of cell groups 6c and 6d for control of a circulating current, each cell 1 can output a voltage of both polarities and hence proportional gain K can also be set to 0.

Operation of Gate Control Portions 5n and 5o

Gate control portion 5n provides corresponding gate signals Gp2u, Gp2v, and Gp2w to the switching element of cell 1 constituting positive-side cell group 6c of the corresponding phase based on voltage command values Vpref2u, Vpref2,v, and Vpref2w of the U phase, the V phase, and the W phase output from an multiplier 5b20. Gate control portion 5o provides corresponding gate signals Gn2u, Gn2v, and Gn2w to the switching element of cell 1 constituting negative-side cell group 6d of each phase based on voltage command values Vnref2,u, Vnref2,v, and Vnref2w of the U phase, the V phase, and the W phase output from multiplier 5b21. Gate control portions 5n and 5o can be operated under a pulse width modulation scheme similarly to gate control portions 5k and 5m.

Effect of First Embodiment

As set forth above, the power conversion device according to the first embodiment includes cell groups 6a and 6b which exclusively (that is, without being used for control of a circulating current) control an electric quantity (a current and a voltage) of each of AC terminals Nu, Nv, and Nw and DC terminals Np and Nn, which is a main purpose of the power conversion device. With cell groups 6a and 6b, an electric quantity of each of
AC terminals Nu, Nv, and Nw and DC terminals Np and Nn can reliably be controlled without interference by control of a circulating current.

Furthermore, the power conversion device according to the first embodiment can control a value of a circulating current in accordance with a circulating current command value by including cell groups 6c and 6d for control of a circulating current. Voltage command value Vpref2 for positive-side cell group 6c is generated by multiplying a value in proportion to voltage command value Vpref for positive-side cell group 6a by voltage command value Vccref for control of a circulating current. Similarly, voltage command value Vpref2 for negative-side cell group 6d is generated by multiplying a value in proportion to voltage command value Vnref for negative-side cell group 6b by voltage command value Vccref for control of a circulating current. One of cell groups 6c and 6d thus mainly controls a circulating current, the other mainly controls a voltage of DC capacitor 1e to be constant, and switching between roles in control is made for each half cycle of the alternating current. Consequently, control of the circulating current and maintaining of a voltage of the DC capacitor can reliably both be achieved.

Modification

In the embodiment above, an example in which each cell 1 constituting cell groups 6a and 6b which are not for control of a circulating current and each cell 1 constituting cell groups 6c and 6d for control of a circulating current are identical in configuration is shown. Unlike this configuration, each cell constituting cell groups 6a and 6b and each cell constituting cell groups 6c and 6d may be different in configuration from each other. An effect the same as above is achieved also in this case.

In each leg circuit 8 in FIG. 1, cell groups 6c and 6d alone can also be provided without providing cell groups 6a and 6b. In this case, gate control portions 5k and 5m in FIG. 3 are not required either. According to such a configuration as well, circulating current control portion 5b generates voltage command values Vpref2 and Vnref2 for output voltages from cell groups 6c and 6d through a non-linear operation (specifically, multiplication) of voltage command values Vpref and Vnref for control of electric quantities of AC terminals Nu, Nv, and Nw and electric quantities of DC terminals Np and Nn and voltage command value Vccref for control of a circulating current. When voltage command values Vpref2 and Vnref2 for cell groups 6c and 6d are generated through linear combination between voltage command values Vpref and Vnref and voltage command value Vccref, control of a circulating current and control of electric quantities of the AC terminal and the DC terminal may interfere with each other. In contrast, by generating voltage command values Vpref2 and Vnref2 for cell groups 6c and 6d only through a non-linear operation of voltage command values Vpref and Vnref and voltage command value
Vccref (that is, without using linear combination), control of a circulating current and control of electric quantities of the AC terminal and the DC terminal can both be achieved.

When cell groups 6a and 6b are not provided (or attention is paid only to cell groups 6c and 6d), control device 5 performs a non-linear operation of a voltage command value for suppression of a circulating current which circulates through leg circuit 8 and at least one voltage command value for control of other currents (a DC current and an AC current) which flow through leg circuit 8. It can then be assumed that a plurality of chopper cells 20 constituting each leg circuit 8 operate in accordance with a result of the non-linear operation.

Second Embodiment

FIG. 6 is a diagram showing a configuration of circulating current control portion 5b in the power conversion device according to a second embodiment. Since the overall configuration of the power conversion device is the same as described with reference to FIG. 1 and the configuration of each cell 1 is the same as described with reference to FIG. 2, description will not be repeated. Since the configuration of control device 5 is the same as described with reference to FIG. 3 except for the configuration of circulating current control portion 5b, description will not be repeated.

Overview of Configuration and Operation of Circulating Current Control Portion 5b

Referring to FIG. 6, circulating current control portion 5b in the second embodiment includes subtractor 5b9, compensator 5b18, a non-linear transfer function application portion 5b22, and adders 5b23 and 5b24.

Subtractor 5b9 calculates a difference between circulating current command value Iccref (Iccrefu for the U phase, Iccrefv for the V phase, and Iccrefw for the W phase) and circulating current value Icc (Iccu of the U phase, Iccv of the V phase, and Iccw of the W phase) by subtracting the circulating current value from the circulating current command value. Compensator 5b18 generates voltage command value Vccref for control of a circulating current (Vccrefu for the U phase, Vccrefv for the V phase, and Vccrefw for the W phase) by amplifying the calculated difference.

An output value (voltage command value Vccref) from compensator 5b18 is subjected to a non-linear operation by non-linear transfer function application portion 5b22. Adder 5b23 generates voltage command value Vpref2 for positive-side cell group 6c by adding a result of operation by non-linear transfer function application portion 5b22 to positive-side voltage correction value Vpcorr. Adder 5b24 generates voltage command value Vnref2 for negative-side cell group 6d by adding a result of operation by non-linear transfer function application portion 5b22 to negative-side voltage correction value Vncorr. As described in the first embodiment, voltage correction values Vpcorr and Vncorr are in proportion to voltage command values Vpref and Vnref for cell groups 6a and 6b, respectively.

Details of Operation of Circulating Current Control Portion 5b

A detailed operation of circulating current control portion 5b will now be described. Circulating current values Iccu, Iccv, and Iccw of the U phase, the V phase, and the W phase operated by current operation portion 5a are sent to circulating current control portion 5b. Circulating current control portion 5b subjects the circulating current value to feedback control such that the circulating current value matches with the circulating current command value. Therefore, circulating current control portion 5b is provided with compensator 5b18 which amplifies a difference between a circulating current command value and a circulating current value. Non-linear transfer function application portion 5b22 is applied to an output value from compensator 5b18 (voltage command value Vccref) (subjects the output value to a non-linear operation).

Non-linear transfer function application portion 5b22 operates to increase sensitivity (increase a ratio of an output to an input) when magnitude of the input (absolute value) is not smaller than a threshold value and to lower sensitivity (lower a ratio of an output to an input) when an absolute value of the input is smaller than the threshold value. For example, non-linear transfer function application portion 5b22 can be implemented by a table for determining input and output characteristics or determination of a condition.

FIG. 7 is a diagram showing one example of relation between an input to and an output from non-linear transfer function application portion 5b22 in FIG. 6. Referring to FIG. 7, an input to non-linear transfer function application portion 5b22 is denoted by x and an output therefrom is denoted by y. An absolute value of input x is denoted by ABS(x), a sign of input x is denoted by SIGN(x), and K1, K2, and A are defined as constants. Then, output y is given in expressions below.

When $ABS (x) < A, y = K 1 \cdot x$

When $ABS (x) \geq A, y = K 2 \cdot x - SIGN (x) \cdot (K 2 - K 1) \cdot A$

In the expressions above, by setting constants K1 and K2 so as to satisfy a condition of K1 < K2, a gain by which an output from compensator 5b18 (input x) is multiplied increases from K1 to K2 when the output from compensator 5b18 is not smaller than constant A. Consequently, non-linear transfer function application portion 5b22 outputs a greater voltage command value (output y) for control of a circulating current.

As non-linear transfer function application portion 5b22 thus operates, greater voltage command value Vccref is output as a difference between circulating current Ice and a command value thereof is greater. Therefore, higher importance is placed on control of a circulating current (voltage command value Vccref) than on a function to maintain a voltage of the DC capacitor (voltage correction values Vpcorr and Vncorr). In contrast, smaller voltage command value Vccref is output as a difference between circulating current Ice and a command value thereof is smaller. Therefore, higher importance is placed on the function to maintain a voltage of the DC capacitor (voltage correction values Vpcorr and Vncorr) than on control of a circulating current (voltage command value Vccref). Consequently, control of a circulating current and maintaining of a voltage of the DC capacitor can both be achieved.

Effect of Second Embodiment

The power conversion device according to the second embodiment includes cell groups 6a and 6b which exclusively (that is, without being used for control of a circulating current) control an electric quantity (a current and a voltage) of each of AC terminals Nu, Nv, and Nw and DC terminals Np and Nn, which is a main purpose of the power conversion device as in the first embodiment. With cell groups 6a and 6b, an electric quantity of each of AC terminals Nu, Nv, and Nw and DC terminals Np and Nn can reliably be controlled without interference by control of a circulating current.

The power conversion device according to the second embodiment can control a circulating current value in accordance with a circulating current command value by including cell groups 6c an 6d for control of a circulating current. Voltage command value Vpref2 for positive-side cell group 6c is generated by adding a value in proportion to voltage command value Vpref for positive-side cell group 6a after a non-linear operation of voltage command value Vccref for control of a circulating current by non-linear transfer function application portion 5b22. Similarly, voltage command value Vnref2 for negative-side cell group 6d is generated by adding a value in proportion to voltage command value Vnref for negative-side cell group 6b after a non-linear operation of voltage command value Vccref for control of a circulating current by non-linear transfer function application portion 5b22. As a difference between circulating current Ice and a command value thereof is thus greater, a function to control a circulating current is exhibited more than a function to maintain a voltage of the DC capacitor. Therefore, control of a circulating current and maintaining of a voltage of the DC capacitor can reliably both be achieved.

Modification

In each leg circuit 8, only reactor 7a on the positive side of reactors 7a and 7b may be provided or only reactor 7b on the negative side may be provided. When only reactor 7b on the negative side is provided, positive-side cell group 6c for control of a circulating current is not required and gate control portion 5n, adder 5b23, and gain circuit 5g associated therewith are not required either, which is advantageous in simplification of the configuration of control device 5. Similarly, when only reactor 7a on the positive side is provided, negative-side cell group 6d for control of a circulating current is not required and gate control portion 5o, adder 5b24, and gain circuit 5h associated therewith are not required either, which is advantageous in simplification of the configuration of control device 5.

Though an example in which a non-linear transfer function (5b22) has three constants is shown above, a high-order transfer function having more constants may be applicable.

In each leg circuit 8 in FIG. 1, cell groups 6c and 6d alone can also be provided without providing cell groups 6a and 6b. In this case, gate control portions 5k and 5m in FIG. 3 are not required either. According to such a configuration, circulating current control portion 5b generates voltage command values Vpref2 and Vnref2 for controlling output voltages from cell groups 6c and 6d through a non-linear operation (specifically by non-linear transfer function application portion 5b22) of voltage command values Vpref and Vnref for control of electric quantities of AC terminals Nu, Nv, and Nw and electric quantities of DC terminals Np and Nn and voltage command value Vccref for control of a circulating current. When voltage command values Vpref2 and Vnref2 for cell groups 6c and 6d are generated through linear combination between voltage command values Vpref and Vnref and voltage command value Vccref, control of a circulating current and control of electric quantities of the AC terminal and the DC terminal may interfere with each other. In contrast, by generating voltage command values Vpref2 and Vnref2 for cell groups 6c and 6d only through a non-linear operation of voltage command values Vpref and Vnref and voltage command value Vccref (that is, without using linear combination), control of a circulating current and control of electric quantities of the AC terminal and the DC terminal can both be achieved.

Third Embodiment

FIG. 8 is a diagram showing a configuration of circulating current control portion 5b in the power conversion device according to a third embodiment. Since the overall configuration of the power conversion device is the same as described with reference to FIG. 1 and the configuration of each cell 1 is the same as described with reference to FIG. 2, description will not be repeated. Since the configuration of control device 5 is the same as described with reference to FIG. 3 except for the configuration of circulating current control portion 5b, description will not be repeated.

Configuration and Operation of Circulating Current Control Portion 5b

Circulating current control portion 5b in FIG. 8 is different from circulating current control portion 5b in FIG. 6 in that a non-linear mathematical function application portion 5b25 is employed instead of non-linear transfer function application portion 5b22. Since FIG. 8 is otherwise the same as FIG. 6, the same or corresponding elements have the same reference characters allotted and description will not be repeated.

In this embodiment, non-linear mathematical function application portion 5b25 applies a smooth non-linear mathematical function to an output from compensator 5b18 (voltage command value Vccref for control of a circulating current). An input to non-linear mathematical function application portion 5b25 is denoted by x, an output therefrom is denoted by y, a hyperbolic sine function of input x is denoted as sinh(x), and A and B are defined as constants. Output y is then given in an expression below. $Y = B \cdot \sinh (A \cdot x)$

The function expressed in the expression (13) is smaller in differential coefficient with decrease in magnitude of the output from compensator 5b18 (input x) (a ratio of an output to an input is lower), and the differential coefficient is greater (a ratio of an output to an input is greater) with increase in magnitude of an output from compensator 5b18. Constants A and B are used for adjustment of input and output characteristics of non-linear mathematical function application portion 5b25.

As non-linear mathematical function application portion 5b25 thus operates, greater voltage command value Vccref is output as a difference between circulating current Icc and a command value thereof is greater. Therefore, higher importance is placed on control of a circulating current (voltage command value Vccref) than the function to maintain a voltage of the DC capacitor (voltage correction values Vpcorr and Vncorr). In contrast, smaller voltage command value Vccref is output as a difference between circulating current Ice and a command value thereof is smaller. Therefore, higher importance is placed on the function to maintain a voltage of the DC capacitor (voltage correction values Vpcorr and Vncorr) than on control of a circulating current (voltage command value Vccref). Consequently, control of a circulating current and maintaining of a voltage of the DC capacitor can both be achieved.

By using a non-linear mathematical function (5b25) of which output y is smoothly varied in response to variation in input x, a discontinuous operation of circulating current control portion 5b can be suppressed and hence feedback control of a circulating current is effectively stabilized.

Effect of Third Embodiment

The power conversion device according to the third embodiment includes cell groups 6a and 6b which exclusively (that is, without being used for control of a circulating current) control an electric quantity (a current and a voltage) of each of AC terminals Nu, Nv, and Nw and DC terminals Np and Nn, which is a main purpose of the power conversion device, as in the first embodiment. With cell groups 6a and 6b, an electric quantity of each of AC terminals Nu, Nv, and Nw and DC terminals Np and Nn can reliably be controlled without interference by control of a circulating current.

The power conversion device according to the third embodiment can control a circulating current value in accordance with a circulating current command value by including cell groups 6c and 6d for control of a circulating current. Voltage command value Vpref2 for positive-side cell group 6c is generated by addition of a value in proportion to voltage command value Vpref for positive-side cell group 6a after a non-linear operation of voltage command value Vccref for control of a circulating current by non-linear mathematical function application portion 5b25. Similarly, voltage command value Vnref2 for negative-side cell group 6d is generated by addition of a value in proportion to voltage command value Vnref for negative-side cell group 6b after a non-linear operation of voltage command value Vccref for control of a circulating current by non-linear function mathematical portion 5b25. As a difference between circulating current Ice and a command value thereof is thus greater, the function to control a circulating current is exhibited more than the function to maintain a voltage of the DC capacitor and therefore control of the circulating current and maintaining of a voltage of the DC capacitor can reliably both be achieved.

Modification

In the embodiment above, an example in which each cell 1 constituting cell groups 6a and 6b which are not for control of a circulating current and each cell 1 constituting cell groups 6c and 6d for control of a circulating current are identical in configuration is shown.
Unlike this configuration, each cell constituting cell groups 6a and 6b and each cell constituting cell groups 6c and 6d may be different in configuration from each other. An effect the same as above is achieved also in this case.

A non-linear mathematical function (5b25) is not limited to the example above, and the effect the same as above is achieved so long as a non-linear mathematical function has such characteristics that a differential coefficient is greater with increase in input x.

In each leg circuit 8 in FIG. 1, cell groups 6c and 6d alone can also be provided without providing cell groups 6a and 6b. In this case, gate control portions 5k and 5m in FIG. 3 are not required either. According to such a configuration, circulating current control portion 5b generates voltage command values Vpref2 and Vnref2 for controlling output voltages from cell groups 6c and 6d through a non-linear operation (specifically by non-linear mathematical function application portion 5b25) of voltage command values Vpref and Vnref for controlling electric quantities of AC terminals Nu, Nv, and Nw and electric quantities of DC terminals Np and Nn and voltage command value Vccref for control of a circulating current. When voltage command values Vpref2 and Vnref2 for cell groups 6c and 6d are generated through linear combination between voltage command values Vpref and Vnref and voltage command value Vccref, control of a circulating current and control of the electric quantities of the AC terminal and the DC terminal may interfere with each other. In contrast, by generating voltage command values Vpref2 and Vnref2 for cell groups 6c and 6d only through a non-linear operation of voltage command values Vpref and Vnref and voltage command value Vccref (that is, without using linear combination), control of a circulating current and control of the electric quantities of the AC terminal and the DC terminal can both be achieved.

Fourth Embodiment

FIG. 9 is a diagram showing a configuration of circulating current control portion 5b in the power conversion device according to a fourth embodiment. Since the overall configuration of the power conversion device is the same as described with reference to FIG. 1 and the configuration of each cell 1 is the same as described with reference to FIG. 2, description will not be repeated. Since the configuration of control device 5 is the same as described with reference to FIG. 3 except for the configuration of circulating current control portion 5b, description will not be repeated.

Configuration and Operation of Circulating Current Control Portion 5b

Referring to FIG. 9, circulating current control portion 5b includes subtractor 5b9, compensator 5b18, a non-linear mathematical function application portion 5b26, and a non-linear mathematical function application portion 5b27. Since subtractor 5b9 and compensator 5b18 are the same as described with reference to FIGS. 6 and 8, description will not be repeated.

Non-linear mathematical function application portion 5b26 receives positive-side voltage correction value Vpcorr and an output value (voltage command value Vccref) from compensator 5b18 as inputs x and y and outputs a value z obtained by applying a smooth non-linear function to inputs x and y as voltage command value Vpref2 for positive-side cell group 6c. Similarly, non-linear mathematical function application portion 5b27 receives negative-side voltage correction value Vncorr and an output value from compensator 5b18 (voltage command value Vccref) as inputs x and y and outputs a value z obtained by applying a smooth non-linear function to inputs x and y as voltage command value Vnref2 for negative-side cell group 6d.

As described already, positive-side voltage correction value Vpcorr is a command value for maintaining a constant capacitor voltage of each cell in positive-side cell group 6c. Negative-side voltage correction value Vncorr is a command value for maintaining a constant capacitor voltage of each cell in negative-side cell group 6d. Voltage command value Vccref output from compensator 5b18 is a command value for control of a circulating current.

Non-linear mathematical function application portions 5b26 and 5b27 apply the same non-linear function to input x (positive-side voltage correction value Vpcorr or negative-side voltage correction value Vncorr) and input y (voltage command value Vccref) in the present embodiment. Specifically, output z from each of non-linear mathematical function application portions 5b26 and 5b27 is given in an expression (14) below with A and B being defined as constants. $z = B \cdot x / [1 + \exp (- A \cdot (1 - x) \cdot y)]$

The function expressed in the expression (14) is smaller in its function value (output z) with decrease in magnitude of an output (input y) from compensator 5b18 and is greater in its function value (output z) with increase in magnitude of an output (input y) from compensator 5b18. The function value (output z) is greater with increase in voltage correction value Vpcorr or Vncorr (input x) and the function value (output z) is smaller with decrease in voltage correction value (input x). Constants A and B (both of which are positive) are used for adjustment of input and output characteristics of non-linear mathematical function application portions 5b26 and 5b27.

FIG. 10 shows one example of the function expressed in the expression (14). The graph in FIG. 10 shows values of output z for a range of x from 0 to 1 and a range of y from -0.2 to 0.2 when constant A is set to 8.47 and constant B is set to 2. The function expressed in the expression (14) has such characteristics that, with input x being around 0, output z is greater than 0 even when y is varied, and with input x being around 1, output z is smaller than 1 even when y is varied.

With such an operation by non-linear mathematical function application portions 5b26 and 5b27, voltage command value Vccref is generated in accordance with a difference between circulating current Icc and a command value thereof, and voltage command values Vpref2 and Vnref2 are output in accordance with voltage command value Vccref and voltage correction values Vpcorr and Vncorr. Therefore, the function to control a circulating current (voltage command value Vccref) and the function to maintain a voltage of the DC capacitor (voltage correction values Vpcorr and Vncorr) simultaneously operate and voltage command values Vpref2 and Vnref2 can smoothly be restricted within a range of a duty in which the chopper cells can operate. Consequently, control of a circulating current and maintaining of a voltage of the DC capacitor can both be achieved without saturation of outputs from the chopper cells.

Since a discontinuous operation of circulating current control portion 5b can be suppressed by using a non-linear mathematical function (5b26 and 5b27) of which output z is smoothly varied in response to variation in inputs x and y, feedback control of a circulating current is effectively stabilized.

Effect of Fourth Embodiment

The power conversion device according to the fourth embodiment includes cell groups 6a and 6b which exclusively (that is, without being used for control of a circulating current) control an electric quantity (a current and a voltage) of each of AC terminals Nu, Nv, and Nw and DC terminals Np and Nn, which is a main purpose of the power conversion device, as in the first embodiment. With cell groups 6a and 6b, an electric quantity of each of AC terminals Nu, Nv, and Nw and DC terminals Np and Nn can reliably be controlled without interference by control of a circulating current.

The power conversion device according to the fourth embodiment can control a circulating current value in accordance with a circulating current command value by including cell groups 6c and 6d for control of a circulating current. Voltage command value Vpref2 for positive-side cell group 6c is generated by subjecting voltage command value Vccref for control of a circulating current and voltage correction value Vpcorr to a non-linear operation by non-linear mathematical function application portion 5b26. Similarly, voltage command value Vnref2 for negative-side cell group 6d is generated by subjecting voltage command value Vccref for control of a circulating current and voltage correction value Vncorr to a non-linear operation by non-linear mathematical function application portion 5b27. Since voltage command values Vpref2 and Vnref2 thus operate without exceeding a range of duties of the chopper cells, control of a circulating current and maintaining of a voltage of the DC capacitor can reliably both be achieved.

Modification

In each leg circuit 8, only reactor 7a on the positive side of reactors 7a and 7b may be provided or only reactor 7b on the negative side may be provided. When only reactor 7b on the negative side is provided, positive-side cell group 6c for control of a circulating current is not required and gate control portion 5n, non-linear mathematical function application portion 5b26, and gain circuit 5g associated therewith are not required either, which is advantageous in simplification of the configuration of control device 5. Similarly, when only reactor 7a on the positive side is provided, negative-side cell group 6d for control of a circulating current is not required and gate control portion 5o, non-linear mathematical function application portion 5b27, and gain circuit 5h associated therewith are not required either, which is advantageous in simplification of the configuration of control device 5.

A non-linear mathematical function (5b26 and 5b27) is not limited to the example above, and the effect the same as above is achieved so long as a non-linear mathematical function has such characteristics that a function value (output z) thereof is greater with increase in inputs x and y and a function value thereof (output z) is within a determined range (in the example above, within a range from 0 to 1).

In each leg circuit 8 in FIG. 1, cell groups 6c and 6d alone can also be provided without providing cell groups 6a and 6b. In this case, gate control portions 5k and 5m in FIG. 3 are not required either. According to such a configuration, circulating current control portion 5b generates voltage command values Vpref2 and Vnref2 for controlling output voltages from cell groups 6c and 6d through a non-linear operation (specifically by non-linear mathematical function application portions 5b26 and 5b27) of voltage command values Vpref and Vnref for controlling electric quantities of AC terminals Nu, Nv, and Nw and electric quantities of DC terminals Np and Nn and voltage command value Vccref for control of a circulating current. When voltage command values Vpref2 and Vnref2 for cell groups 6c and 6d are generated by linear combination between voltage command values Vpref and Vnref and voltage command value Vccref, control of a circulating current and control of the electric quantities of the AC terminal and the DC terminal may interfere with each other. In contrast, by generating voltage command values Vpref2 and Vnref2 for cell groups 6c and 6d only through a non-linear operation of voltage command values Vpref and Vnref and voltage command value Vccref (that is, without using linear combination), control of a circulating current and control of the electric quantities of AC terminals Nu, Nv, and Nw and DC terminals Np and Nn can both be achieved.

Fifth Embodiment

FIG. 11 is a diagram showing a configuration of circulating current control portion 5b in the power conversion device according to a fifth embodiment. Since the overall configuration of the power conversion device is the same as described with reference to FIG. 1 and the configuration of each cell 1 is the same as described with reference to FIG. 2, description will not be repeated. Since the configuration of control device 5 is the same as described with reference to FIG. 3 except for the configuration of circulating current control portion 5b, description will not be repeated.

Overview of Configuration and Operation of Circulating Current Control Portion 5b

Referring to FIG. 11, in the fifth embodiment, circulating current control portion 5b includes adders 5b1, 5b3, 5b5, and 5b7, subtractors 5b2, 5b4, 5b6, 5b8, and 5b9, compensators 5b18p and 5b18n, and limiters 5b13 and 5b17. Compensator 5b18p is a proportional integral compensator and includes a gain circuit 5b10, an integrator 5b11 with a limiter, and an adder 5b12. Similarly, compensator 5b18n is a proportional integral compensator and includes a gain circuit 5b14, an integrator 5b15 with a limiter, and an adder 5b16.

Compensator 5b18p generates a voltage command value Vpccref for control of a circulating current (Vpccrefu for the U phase, Vpccrefv for the V phase, and Vpccrefw for the W phase) by amplifying the difference calculated by subtractor 5b9. In compensator 5b18p, gain circuit 5b10 amplifies the difference output from subtractor 5b9. Integrator 5b11 integrates the output from gain circuit 5b10. Adder 5b12 adds the output from gain circuit 5b10 and the output from integrator 5b11 to each other. The output from adder 5b12 is used as voltage command value Vpref2 for positive-side cell group 6c after it passes through limiter 5b13.

Similarly, compensator 5b18n generates a voltage command value Vnccref for control of a circulating current (Vnccrefu for the U phase, Vnccrefv for the V phase, and Vnccrefw for the W phase) by amplifying the difference calculated by subtractor 5b9. In compensator 5b18n, gain circuit 5b14 amplifies the difference output from subtractor 5b9. Integrator 5b15 integrates the output from gain circuit 5b14. Adder 5b16 adds the output from gain circuit 5b14 and the output from integrator 5b15 to each other. The output from adder 5b16 is used as voltage command value Vnref2 for negative-side cell group 6d after it passes through limiter 5b17.

Voltage correction values Vpcorr and Vncorr in addition to circulating current Icc are input to circulating current control portion 5b. As described in the first embodiment, voltage correction values Vpcorr and Vncorr are in proportion to voltage command values Vpref and Vnref for cell groups 6a and 6b, respectively.

Adder 5b 1 generates an upper limit value of limiter 5b 13 by adding a fixed value Δ to voltage correction value Vpcorr. Subtractor 5b2 generates a lower limit value of limiter 5b13 by subtracting fixed value Δ from voltage correction value Vpcorr. Adder 5b3 generates an output upper limit value of integrator 5b11 by adding a fixed value Δi to voltage correction value Vpcorr. Subtractor 5b4 generates an output lower limit value of integrator 5b11 by subtracting fixed value Δi from voltage correction value Vpcorr. In order to prevent wind-up phenomena, fixed value Δi should be smaller than fixed value Δ.

Similarly, adder 5b5 generates an upper limit value of limiter 5b17 by adding fixed value Δ to voltage correction value Vncorr. Subtractor 5b6 generates a lower limit value of limiter 5b17 by subtracting fixed value Δ from voltage correction value Vncorr. Adder 5b7 generates an output upper limit value of integrator 5b 15 by adding fixed value Δi to voltage correction value Vncorr. Subtractor 5b8 generates an output lower limit value of integrator 5b15 by subtracting fixed value Δi from voltage correction value Vncorr. In order to prevent wind-up phenomena, fixed value Δi should be smaller than fixed value Δ.

Details of Operation of Circulating Current Control Portion 5b

A detailed operation of circulating current control portion 5b will now be described. Though generation of voltage command value Vpref2 for positive-side cell group 6c is mainly described below, similar description is applicable also to generation of voltage command value Vnref2 for negative-side cell group 6d.

Circulating current value Icc (Iccu of the U phase, Iccv of the V phase, and Iccw of the W phase) operated by current operation portion 5a is sent to circulating current control portion 5b. Circulating current control portion 5b subjects the circulating current value to feedback control such that the circulating current value matches with a circulating current command value.

Specifically, initially, subtractor 5b9 calculates a difference between circulating current value Icc and circulating current command value Iccref. The difference between circulating current Icc and command value Iccref thereof is amplified (multiplied by a proportional constant) by gain circuit 5b10 and subjected to time integration by integrator 5b11. Upper and lower limit values of the output from integrator 5b11 are restricted based on positive-side voltage correction value Vpcorr, and a value within a range of Vpcorr±Δi is output.

FIG. 12 is a diagram schematically showing a waveform showing change over time of voltage command value Vpref2 output from circulating current control portion 5b. As shown in FIG. 12, circulating current control portion 5b outputs a value within a range of Vpcorr±Δi (a hatched region in FIG. 12) as voltage command value Vpref2.

When the difference between circulating current Icc and command value Iccref thereof is zero, an output from gain circuit 5b10 is also zero and hence adder 5b12 outputs a value around positive-side voltage correction value Vpcorr. With this operation, active power output from cell group 6c or input to cell group 6c is 0, and therefore a voltage of DC capacitor 1e in each cell 1 constituting cell group 6c is maintained at a constant value.

When the difference between circulating current Icc and command value Iccref thereof increases, gain circuit 5b10 outputs a value obtained by amplifying the difference so that circulating current control portion 5b operates to suppress the difference. In this operation, an output from compensator 5b18p is restricted by Vpcorr±Δ. By preventing the wind-up phenomena by setting a condition of Δ > Δi, the output from gain circuit 5b10 is reliably reflected on voltage command value Vpref2 for positive-side cell group 6c. The difference in circulating current can thus be suppressed. By including integrator 5b11 in compensator 5b18p, variation in voltage caused by a resistive loss in reactors 7a and 7b can be compensated for.

Effect of Fifth Embodiment

The power conversion device according to the fifth embodiment includes cell groups 6a and 6b which exclusively (that is, without being used for control of a circulating current) control an electric quantity (a current and a voltage) of each of AC terminals Nu, Nv, and Nw and DC terminals Np and Nn, which is a main purpose of the power conversion device, as in the first embodiment. With cell groups 6a and 6b, an electric quantity of each of AC terminals Nu, Nv, and Nw and DC terminals Np and Nn can reliably be controlled without interference by control of a circulating current.

The power conversion device according to the fifth embodiment can control a circulating current value in accordance with a circulating current command value by including cell groups 6c and 6d for control of a circulating current. Circulating current control portion 5b performs an operation on a difference between circulating current Ice and command value Iccref thereof by using proportional integral compensators 5b18p and 5b18n and outputs from compensators 5b18p and 5b18n are restricted by limiters 5b13 and 5b17 to be around voltage correction values Vpcorr and Vncorr. By controlling cell groups 6c and 6d with voltage command values Vpref2 and Vnref2 thus generated, a circulating current can be controlled while a voltage of the DC capacitor in each cell is maintained at a constant value. Integrators 5b11 and 5b15 constituting compensators 5b18p and 5b18n can suppress variation in voltage due to a resistive loss in reactors 7a and 7b.

Modification

In each leg circuit 8, only reactor 7a on the positive side of reactors 7a and 7b may be provided or only reactor 7b on the negative side may be provided. When only reactor 7b on the negative side is provided, positive-side cell group 6c for control of a circulating current is not required and gate control portion 5n, adders 5b1 and 5b3, subtractors 5b2 and 5b4, compensator 5b18p, and limiter 5b13 associated therewith are not required either, which is advantageous in simplification of the configuration of control device 5. Similarly, when only reactor 7a on the positive side is provided, negative-side cell group 6d for control of a circulating current is not required and gate control portion 5o, adders 5b5 and 5b7, subtractors 5b6 and 5b8, compensator 5bl8n, and limiter 5b17 associated therewith are not required either, which is advantageous in simplification of the configuration of
control device 5.

In each leg circuit 8 in FIG. 1, cell groups 6c and 6d alone can also be provided without providing cell groups 6a and 6b. In this case, gate control portions 5k and 5m in FIG. 3 are not required either. According to such a configuration, circulating current control portion 5b generates voltage command values Vpref2 and Vnref2 for controlling output voltages from cell groups 6c and 6d through a non-linear operation (specifically by integrator 5b11 with a limiter and limiter 5b13) of voltage command values Vpref and Vnref for controlling electric quantities of AC terminals Nu, Nv, and Nw and electric quantities of DC terminals Np and Nn and voltage command value Vccref for control of a circulating current. When voltage command values Vpref2 and Vnref2 for cell groups 6c and 6d are generated through linear combination between voltage command values Vpref and Vnref and voltage command value Vccref, control of a circulating current and control of the electric quantities of the AC terminal and the DC terminal may interfere with each other. In contrast, by generating voltage command values Vpref2 and Vnref2 for cell groups 6c and 6d only through a non-linear operation of voltage command values Vpref and Vnref and voltage command value Vccref (that is, without using linear combination), control of a circulating current and control of the electric quantities of the AC terminal and the DC terminal can both be achieved.

Sixth Embodiment Configuration of Power Conversion Device

FIG. 13 is a schematic configuration diagram of a power conversion device according to a sixth embodiment. The power conversion device in FIG. 13 is different from the power conversion device in FIG. 1 in configuration of each cell 20 provided in cell groups 6c and 6d for control of a circulating current. Specifically, each converter cell 20 provided in cell groups 6c and 6d in FIG. 13 is configured to detect a voltage of DC capacitor 1e (which is hereinafter referred to as a cell capacitor voltage Vccell) provided in the converter cell itself and to transmit a detection value to control device 5. Since FIG. 13 is otherwise the same in configuration as FIG. 1, description will not be repeated.

FIG. 14 is a circuit diagram showing a detailed configuration of each cell 20 provided in cell groups 6c and 6d for control of a circulating current. FIG. 14 shows exemplary cell 20 of a half bridge type.

Referring to FIG. 14, converter cell 20 is different from converter cell 1 in FIG. 2(a) in further including a DC voltage detector 1j provided in parallel to DC capacitor 1e. DC voltage detector 1j detects voltage Vccell of DC capacitor 1e and outputs detected
cell capacitor voltage Vccell to control device 5.

Converter cell 20 may be configured as a full bridge in FIG. 2(b) or may be configured with the configuration in FIG. 2(c) being made use of. DC voltage detector 1j is provided in parallel to DC capacitor 1e also in these examples.

Configuration of Control Device 5

FIG. 15 is a configuration diagram of control device 5 in FIG. 13. Control device 5 shown in FIG. 15 is different from control device 5 in FIG. 3 in further including a voltage operation portion 5p and capacitor voltage control portions 5q and 5r. Since FIG. 15 is otherwise the same in configuration as FIG. 13, elements the same as those or elements corresponding to those in FIG. 13 have the same reference characters allotted and description may not be repeated below.

Voltage operation portion 5p receives information on cell capacitor voltage Vccell from each cell 20 provided in cell groups 6c and 6d of leg circuits 8a, 8b, and 8c of the respective phases shown in FIG. 13. Voltage operation portion 5p calculates a representative value Vcp (Vcpu of the U phase, Vcpv of the V phase, and Vcpw of the W phase) of a plurality of cell capacitor voltages of positive-side cell group 6c for each of the U phase, the V phase, and the W phase based on received information on cell capacitor voltage Vccell and calculates a representative value Vcn (Vcnu of the U phase, Vcnv of the V phase, and Vcnw of the W phase) of a plurality of cell capacitor voltages of negative-side cell group 6d. An average value, a median, a maximum value, or a minimum value of cell capacitor voltages Vccell of each cell group can be applied as appropriate to operation of the representative value. Voltage operation portion 5p outputs representative values Vcpu, Vcpv, and Vcpw of the cell capacitor voltages of each positive-side cell group 6c to capacitor voltage control portion 5q and representative values Vcnu, Vcnv, and Vcnw of the cell capacitor voltages of each negative-side cell group 6d to capacitor voltage control portion 5r.

Capacitor voltage control portion 5q receives information on DC current value IDC from current operation portion 5a and receives information on cell capacitor voltage values Vcpu, Vcpv, and Vcpw of positive-side cell group 6c from voltage operation portion 5p. Capacitor voltage control portion 5q generates a voltage correction value Vpcorr2 on which voltage command value Vpref2 for positive-side cell group 6c is based based on the received information and outputs generated voltage correction value Vpcorr2 to adder 5i.

Capacitor voltage control portion 5r receives information on DC current value IDC from current operation portion 5a and receives information on cell capacitor voltage values Vcnu, Vcnv, and Vcnw of negative-side cell group 6d from voltage operation portion 5p. Capacitor voltage control portion 5r generates a voltage correction value Vncorr2 on which voltage command value Vnref2 for negative-side cell group 6d is based based on the received information and outputs generated voltage correction value Vncorr2 to adder 5j.

Adder 5i adds a value in proportion to voltage command value Vpref for positive-side cell group 6a (a value multiplied by gain K by gain circuit 5g) and voltage correction value Vpcorr2 output from capacitor voltage control portion 5q to each other and outputs a result of addition as final voltage correction value Vpcorr to circulating current control portion 5b. Similarly, adder 5j adds a value in proportion to voltage command value Vnref for negative-side cell group 6b (a value multiplied by gain K by gain circuit 5h) and voltage correction value Vncorr2 output from capacitor voltage control portion 5r to each other and outputs a result of addition to circulating current control portion 5b as final voltage correction value Vncorr.

A configuration in any of FIGS. 4, 6, 8, 9, and 11 may be applied to circulating current control portion 5b in the sixth embodiment.

Detailed Operation of Control Device 5

A detailed operation of control device 5 will now be described. Description of an operation in common to that in FIG. 3 in the first embodiment will not be repeated.

Since voltages output from cell groups 6c and 6d for control of a circulating current have a function to control a current which flows through reactors 7a and 7b, power output from cell groups 6c and 6d is substantially reactive power. When active power originating from a loss in reactors 7a and 7b is unignorable, however, active power should be supplied to cell groups 6c and 6d, because the method described in the first embodiment, that is, the method of providing only a value in proportion to voltage command values Vpref and Vnref provided to cell groups 6a and 6b to circulating current control portion 5b as voltage correction values Vpcorr and Vncorr alone, cannot maintain a voltage of DC capacitor 1e of cell groups 6c and 6d at a constant value.

From a point of view above, in the power conversion device in FIGS. 13 and 15, voltage detector 1j detects a voltage of DC capacitor 1e of each cell 20 constituting each of cell groups 6c and 6d. Voltage operation portion 5p operates representative values Vcpu, Vcpv, Vcpw, Vcnu, Vcnv, and Vcnw of cell capacitor voltages Vccell (which are simply referred to as capacitor voltage values for the sake of brevity) of each of cell groups 6c and 6d. A compensator provided in each of capacitor voltage control portions 5q and 5r amplifies a difference between a capacitor voltage command value and a capacitor voltage value (that is, a command value - a voltage value) for each of cell groups 6c and 6d of each phase. Voltage control portions 5q and 5r output a result of multiplication of the amplified difference by a polarity (1 or -1) of DC current value IDC to adders 5i and 5j as voltage correction values Vpcorr2 and Vncorr2.

Adder 5i adds voltage command value Vccref for control of a circulating current, a value in proportion to voltage command value Vpref for cell group 6a, and voltage correction value Vpcorr to one another. A result of addition is supplied to gate control portion 5n as voltage command value Vpref2 for cell group 6c. Adder 5j adds voltage command value Vccref for control of a circulating current, a value in proportion to voltage command value Vnref for cell group 6b, and voltage correction value Vncorr to one another. A result of addition is supplied to gate control portion 5o as voltage command value Vnref2 for cell group 6d.

According to the configuration, (i) when DC current value IDC is positive (polarity = 1) and a capacitor voltage is smaller than a command value thereof, the compensator provided in each of capacitor voltage control portions 5q and 5r outputs a positive signal. Therefore, voltage correction values Vpcorr2 and Vncorr2 obtained by multiplication of the output from the compensator by the polarity (=1) of DC current IDC serves as a signal having a positive DC component. With a signal of this voltage correction value, a period during which switching element 1a in FIG. 5 is conducting is longer and hence a period during which DC current IDC flows into DC capacitor 1e is longer. Consequently, DC capacitor 1e is charged, and therefore a difference between the capacitor voltage command value and a detection value of the capacitor voltage is eliminated.

(ii) When DC current value IDC is positive (polarity = 1) and a capacitor voltage is greater than a command value thereof, the compensator provided in each of capacitor voltage control portions 5q and 5r outputs a negative signal. Therefore, voltage correction values Vpcorr2 and Vncorr2 obtained by multiplication of the output from the compensator by the polarity (=1) of DC current IDC serves as a signal having a negative DC component. With a signal of this voltage correction value, a period during which switching element 1a in FIG. 5 is conducting is shorter and therefore a difference between the capacitor voltage command value and a detection value of the capacitor voltage is eliminated.

(iii) When DC current value IDC is negative (polarity = -1) and a capacitor voltage is smaller than a command value thereof, the compensator provided in each of capacitor voltage control portions 5q and 5r outputs a positive signal. Therefore, voltage correction values Vpcorr2 and Vncorr2 obtained by multiplication of the output from the compensator by the polarity (= -1) of DC current IDC serves as a signal having a negative DC component. With the signal of this voltage correction value, a period during which switching element 1a in FIG. 5 is conducting is shorter and therefore a period during which
DC current IDC flows out of DC capacitor 1e is shorter. Consequently, since a time period of discharging of DC capacitor 1e decreases (charged), a difference between the capacitor voltage command value and a detection value of the capacitor voltage is eliminated.

(iv) When DC voltage value IDC is negative (polarity = -1) and a capacitor voltage is greater than a command value thereof, the compensator provided in each of capacitor voltage control portions 5q and 5r outputs a negative signal. Therefore, voltage correction values Vpcorr2 and Vncorr2 obtained by multiplication of the output from the compensator by the polarity (= -1) of DC current IDC serves as a signal having a positive DC component. With the signal of this voltage correction value, a period during which switching element 1a in FIG. 5 is conducting is longer and therefore a time period of discharging of DC capacitor 1e is longer. Therefore, a difference between the capacitor voltage command value and a detection value of the capacitor voltage is eliminated.

Effect of Sixth Embodiment

The power conversion device according to the sixth embodiment includes cell groups 6a and 6b which exclusively (that is, without being used for control of a circulating current) control an electric quantity (a current and a voltage) of each of AC terminals Nu, Nv, and Nw and DC terminals Np and Nn, which is a main purpose of the power conversion device as in the first embodiment. With cell groups 6a and 6b, an electric quantity of each of AC terminals Nu, Nv, and Nw and DC terminals Np and Nn can reliably be controlled without interference by control of a circulating current.

Furthermore, the power conversion device according to the sixth embodiment can control a value of a circulating current in accordance with a circulating current command value by including cell groups 6c and 6d for control of a circulating current. The voltage command values are generated through a non-linear operation of voltage command values Vpref and Vnref for controlling cell groups 6a and 6b, voltage command value Vccref for control of a circulating current, and voltage correction values Vpcorr2 and Vncorr2 based on a cell capacitor voltage. Consequently, a circulating current can be suppressed while active power input to or output from each cell 20 in cell groups 6c and 6d is set to zero. In particular, since control is based on the cell capacitor voltage, a voltage of the DC capacitor can be maintained constant even though a loss in the reactor and/or variation in electric quantity occur(s).

Modification

As in the first embodiment, in each leg circuit 8, only reactor 7a on the positive side of reactors 7a and 7b may be provided or only reactor 7b on the negative side may be provided. When only reactor 7b on the negative side is provided, positive-side cell group 6c for control of a circulating current is not required and gate control portion 5n, adder 5i, gain circuit 5g, and capacitor voltage control portion 5q associated therewith are not required either, which is advantageous in simplification of the configuration of control device 5. Similarly, when only reactor 7a on the positive side is provided, negative-side cell group 6d for control of a circulating current is not required and gate control portion 5o, adder 5j, gain circuit 5h, and capacitor voltage control portion 5r associated therewith are not required either, which is advantageous in simplification of the configuration of control device 5.

Though an example in which capacitor voltage control portions 5q and 5r multiply an output from the compensator by a polarity of DC current value IDC is shown in the embodiment above, the same effect is achieved also by multiplying the output from the compensator by DC current value IDC itself instead of the polarity of DC current value Idc. When DC control portion 5d carries out feedback control based on a difference between a DC current command value and DC current value Idc, the same effect is achieved also by multiplying the output from the compensator by a DC current command value instead of a polarity of DC current value Idc. The same effect is obtained also by multiplying an output from the compensator of each phase by an AC current value of each phase (Iacu of the U phase, Iacv of the V phase, and Iacw of the W phase) or a polarity thereof instead of a polarity of DC current value IDC in capacitor voltage control portion 5q. The same effect is obtained also by multiplying an output from the compensator of each phase by an AC current value opposite in polarity of each phase (-Iacu of the U phase, -Iacv of the V phase, and -Iacw of the W phase) or a polarity thereof in capacitor voltage control portion 5r.

In each leg circuit 8 in FIG. 13, cell groups 6c and 6d alone can also be provided without providing cell groups 6a and 6b. In this case, gate control portions 5k and 5m in FIG. 15 are not required either. According to such a configuration, circulating current control portion 5b generates voltage command values Vpref2 and Vnref2 for controlling output voltages from cell groups 6c and 6d through a non-linear operation of voltage command values Vpref and Vnref for control of electric quantities of AC terminals Nu, Nv, and Nw and electric quantities of DC terminals Np and Nn, voltage command value Vccref for control of a circulating current, and voltage correction values Vpcorr2 and Vncorr2 based on a cell capacitor voltage. When voltage command values Vpref2 and Vnref2 for cell groups 6c and 6d are generated through linear combination between voltage command values Vpref and Vnref and voltage command value Vccref, control of a circulating current and control of electric quantities of the AC terminal and the DC terminal may interfere with each other. In contrast, by generating voltage command values Vpref2 and Vnref2 for cell groups 6c and 6d only through a non-linear operation of voltage command values Vpref and Vnref and voltage command value Vccref (that is, without using linear combination), control of a circulating current and control of electric quantities of the AC terminal and the DC terminal can both be achieved.

Seventh Embodiment

Though a power conversion device in a seventh embodiment is the same as the sixth embodiment shown in FIG. 13 in overall configuration, control device 5 is different in part from FIG. 15 in the sixth embodiment in configuration and operations. Specific description will be given below with reference to FIGS. 13 and 16.

Configuration of Control Device 5

FIG. 16 is a configuration diagram of control device 5 included in the power conversion device according to the seventh embodiment. Control device 5 in FIG. 16 is different from control device 5 in FIG. 15 in that DC voltage command value Vdcref output from DC control portion 5d is input to each of adders 5i and 5j instead of values in proportion to voltage command values Vpref and Vnref. Control device 5 in FIG. 16 is different from control device 5 in FIG. 15 in that AC current values Iacu, Iacv, and Iacw instead of DC current value IDC are input to capacitor voltage control portion 5q. Control device 5 in Fig. 16 is further different from control device 5 in FIG. 15 in that AC current values -Iacu, -Iacv, and -Iacw opposite in polarity which are obtained by multiplication by -1 by a gain circuit 5s instead of DC current value IDC are input to capacitor voltage control portion 5r. Since FIG. 16 is otherwise the same in configuration as FIG. 15, elements the same as those or corresponding to those in FIG. 15 have the same reference characters allotted and description may not be repeated below.

Operation of Control Device 5

An operation of control device 5 in FIG. 16 will now be described. Description of an operation in common to that in FIG. 3 in the first embodiment and FIG. 15 in the sixth embodiment will not be repeated.

Voltage command value Vccref for control of a circulating current output from circulating current control portion 5b (Vccrefu of the U phase, Vccrefv of the V phase, and Vccrefw of the W phase) is a signal having a polarity of both of positive and negative. Therefore, when converter cell 20 constituting cell groups 6c and 6d is configured as a half bridge as shown in FIG. 2(a) or 14, a bias is required for a voltage command value. In the seventh embodiment, the bias is set to DC voltage command value Vdcref output from DC control portion 5d.

Capacitor voltage control portion 5q generates voltage correction values Vpcorr2u, Vpcorr2v, and Vpcorr2w by amplifying a difference between capacitor voltage values
Vcpu, Vcpv, and Vcpw and a capacitor voltage command value for each phase and multiplying the amplified difference by AC current values Iacu, Iacv, and Iacw, respectively. Similarly, capacitor voltage control portion 5r generates voltage correction values Vncorr2u, Vncorr2v, and Vncorr2w for control of a circulating current by amplifying a difference between capacitor voltage values Vcnu, Vcnv, and Vcnw and a capacitor voltage command value for each phase and multiplying the amplified difference by AC current values -Iacu, -Iacv, and -Iacw opposite in polarity, respectively.

When a DC current flows in cell groups 6c and 6d, active power is generated in each cell 20 constituting cell groups 6c and 6d in accordance with DC voltage command value Vdcref representing a DC value and consequently DC capacitor 1e of each cell 20 is charged or discharges. When a difference is thus produced between the voltage of DC capacitor 1e and the capacitor voltage command value, capacitor voltage control portions 5q and 5r generate voltage correction values Vpcorr2 and Vncorr2 by amplifying the difference and multiplying the difference by an AC current value (or an AC current value opposite in polarity). Voltage correction values Vpcorr2 and Vncorr2 serve for control of each cell 20 in cell groups 6c and 6d so as to output an AC voltage in phase with (or opposite in phase to) the AC current. As each cell 20 generates an AC voltage in accordance with voltage correction values Vpcorr and Vncorr, the generated AC voltage is applied to an AC current which actually flows and hence active power is generated. As AC active power and DC power are balanced, a difference between a voltage value of DC capacitor 1e of each cell 20 and the capacitor voltage command value decreases and the DC capacitor voltage is subjected to feedback control so as to match with the capacitor voltage command value.

Effect of Seventh Embodiment

The power conversion device according to the seventh embodiment includes cell groups 6a and 6b which exclusively (that is, without being used for control of a circulating current) control an electric quantity (a current and a voltage) of each of AC terminals Nu, Nv, and Nw and DC terminals Np and Nn, which is a main purpose of the power conversion device as in the sixth embodiment. With cell groups 6a and 6b, an electric quantity of each of AC terminals Nu, Nv, and Nw and DC terminals Np and Nn can reliably be controlled without interference by control of a circulating current.

Furthermore, the power conversion device according to the seventh embodiment can control a value of a circulating current in accordance with a circulating current command value by including cell groups 6c and 6d for control of a circulating current. The voltage command values are generated through a non-linear operation of DC voltage command value Vdcref for controlling electric quantities of DC terminals Np and Nn, voltage command value Vccref for control of a circulating current, and voltage correction values Vpcorr2 and Vncorr2 based on a cell capacitor voltage. Consequently, a circulating current can be suppressed while active power input to or output from each cell 20 in cell groups 6c and 6d is set to zero. In particular, since control is based on the cell capacitor voltage, a voltage of the DC capacitor can be maintained constant even though a loss in the reactor and/or variation in electric quantity occur(s).

Modification

As in the first embodiment, in each leg circuit 8, only reactor 7a on the positive side of reactors 7a and 7b may be provided or only reactor 7b on the negative side may be provided. When only reactor 7b on the negative side is provided, positive-side cell group 6c for control of a circulating current is not required and gate control portion 5n, adder 5i, and capacitor voltage control portion 5q associated therewith are not required either, which is advantageous in simplification of the configuration of control device 5. Similarly, when only reactor 7a on the positive side is provided, negative-side cell group 6d for control of a circulating current is not required and gate control portion 5o, adder 5j, and capacitor voltage control portion 5r associated therewith are not required either, which is advantageous in simplification of the configuration of control device 5.

In the embodiment, the same effect is achieved also when a signal input to adders 5i and 5j is set to a constant bias value instead of DC voltage command value Vdcref.

It should be understood that the embodiments disclosed herein are illustrative and non-restrictive in every respect. The scope of this invention is defined by the terms of the claims rather than the description above and is intended to include any modifications within the scope and meaning equivalent to the terms of the claims.

REFERENCE SIGNS LIST

1, 20: converter cell
1a, 1b, 1f, 1g: switching element
1c, 1d, 1h, 1i: diode
1e DC: capacitor
1j, 11a, 11b DC: voltage detector
1n, 1p: input and output terminal
2: AC circuit
3: interconnected transformer
4: DC circuit
5: control device
5a: current operation portion
5b: circulating current control portion
5b18, 5b18p, 5b18n: compensator
5b20,5b21: multiplier
5b22: non-linear transfer function application portion
5b25: non-linear mathematical function application portion
5b13, 5b17: limiter
5b11, 5b15: integrator
5c: AC control portion
5d: DC control portion
5e, 5f: command value combination portion
5k, 5m, 5n, 5o: gate control portion
5p: voltage operation portion
5q, 5r: capacitor voltage control portion
5z: voltage command value generation portion
6a, 6c: positive-side cell group
6b, 6d: negative-side cell group
7a, 7b: reactor
8, 8a, 8b, 8c: leg circuit
9a, 9b: arm current detector
10: AC voltage detector
13: positive-side arm
14: negative-side arm
Icc: circulating current
IDC: DC current
IAC: AC current
IDC: DC current value
Inu, Inv, Inw, Ipu, Ipv, Ipw: arm current
K: proportional gain
Nn: negative-side DC terminal
Np: positive-side DC terminal
Nu, Nv, Nw: AC terminal (AC connection portion)
VDC: value of voltage across DC terminals

A power conversion device configured to convert power between a DC circuit (4) and an AC circuit (2), the power conversion device comprising: - a plurality of leg circuits (8a, 8b, 8c) which correspond to respective phases of the AC circuit (2) and are connected in parallel between common first and second DC terminals (Np, Nn), wherein each leg circuit is divided into a first arm (13) on a high potential side and a second arm (14) on a low potential side with an interposed connection portion (Nu, Nv, Nw) which is electrically connectable to a corresponding phase of the AC circuit; each one of the first arm (13) and the second arm (14) including: - a plurality of converter cells (1) cascaded to one another and each including an energy storage (1e), the plurality of converter cells (1) including a plurality of first converter cells (6a, 6b) and a plurality of second converter cells (6c, 6d); and

- at least one inductor (7a, 7b) connected in series to the plurality of converter cells; the power conversion device including:

- a control device (5) configured to control operations of the plurality of converter cells, the control device including:

- an AC control portion (5c) configured to generate, separately for each leg circuit, a first voltage command value (Vacref) representing an AC voltage component to be output from the plurality of first converter cells (6a, 6b) of the leg circuit based on an AC current (Iac) and an AC voltage (Vac) of the AC circuit;

- a DC control portion (5d) configured to generate a second voltage command value (Vdcref) representing a DC voltage component to be output from the plurality of first converter cells (6a, 6b) of each leg circuit based on a DC current (Idc) and a DC voltage (Vdc) of the DC circuit; and

- a circulating current control portion (5b) configured to generate, separately for each leg circuit, a third voltage command value (Vccref) representing a voltage to be output from the plurality of second converter cells (6c, 6d) in order to suppress a circulating current (Icc) for the leg circuit based on the circulating current (Icc) which circulates through the leg circuit,
wherein i) the control device (5) includes a first command value combination portion (5e), wherein the first command value combination portion (5e) is configured to calculate, separately for each leg circuit, a fourth voltage command value (Vpref) based on the difference between the first voltage command value (Vacref) and the second voltage command value (Vdcref); and the control device (5) includes a second command value combination portion (5f) wherein the second command value combination portion (5f) is configured to calculate, separately for each leg circuit, a fifth voltage command value (Vnref) based on the sum of the first voltage command value (Vacref) and the second voltage command value (Vdcref); and wherein a gain circuit (5g) is configured to multiply, separately for each leg circuit, the fourth voltage command value (Vpref) by a gain (K) and to output a positive-side voltage correction value (Vpcorr) to the circulating current control portion (5b), and a gain circuit (5h) is configured to multiply, separately for each leg circuit, the fifth voltage command value (Vnref) by a gain (K) and to output a negative-side voltage correction value (Vncorr) to the circulating current control portion (5b), wherein the circulating current control portion (5b) is configured to generate and output, separately for each leg circuit, a positive-side voltage command value (Vpref2) for operating the plurality of second converter cells (6c) of the first arm (13) and a negative-side voltage command value (Vnref2) for operating the plurality of second converter cells (6d) of the second arm (14) by performing a specific operation with the positive-side voltage correction value (Vpcorr), negative-side voltage correction value (Vncorr), and third voltage command value (Vccref, Vpccref, Vnccref);
or

ii) wherein the control device (5) includes a capacitor voltage control portion (5q, 5r) which is configured to generate, separately for each leg circuit, a first voltage correction value (Vpcorr2) and a second voltage correction value (Vncorr2) based on a value calculated by multiplying a difference between a representative value (Vcpu, Vcpv, Vcpw, Vcnu, Vcnv, Vcnw) of a capacitor voltage of the plurality of second converter cells (6c, 6d) and a command value of the capacitor voltage by an AC current value of the AC circuit through feedback control to decrease the difference; and a first adder (5i) is configured to add, separately for each leg circuit, the second voltage command value (Vdcref) to said first voltage correction value (Vpcorr2) and to output, for each leg circuit, a positive-side voltage correction value (Vpcorr) to the circulating current control portion (5b), and a second adder (5j) is configured to add, separately for each leg circuit, the second voltage command value (Vdcref) to said second voltage correction value (Vncorr2) and to output, for each leg circuit, a negative-side voltage correction value (Vncorr) to the circulating current control portion (5b), wherein the circulating current control portion (5b) is configured to generate and output, separately for each leg circuit, a positive-side voltage command value (Vpref2) for operating the plurality of second converter cells (6c) of the first arm (13) and a negative-side voltage command value (Vnref2) for operating the plurality of second converter cells (6d) of the second arm (14) by performing a specific operation with the positive-side voltage correction value (Vpcorr), negative-side voltage correction value (Vncorr) and third voltage command value (Vccref);

- the specific operation according to either one of above options i) or ii) including any one of: a) multiplying each of the positive-side voltage correction value (Vpcorr) and the negative-side voltage correction value (Vncorr) by a sum of the third voltage command value (Vccref) and a fixed value, to output the positive-side voltage command value (Vpref2) and the negative-side voltage command value (Vnref2) respectively;

b) adding, to each of the positive-side voltage correction value (Vpcorr) and the negative-side voltage correction value (Vncorr), a value calculated by applying a non-linear transfer function (5b22) to the third voltage command value (Vccref), wherein the non-linear transfer function (5b22) is configured such that a ratio of an output to an input increases when the input exceeds a threshold value, to output the positive-side voltage command value (Vpref2) and the negative-side voltage command value (Vnref2) respectively;

c) adding, to each of the positive-side voltage correction value (Vpcorr) and the negative-side voltage correction value (Vncorr), a value obtained by applying a non-linear mathematical function (5b25) to the third voltage command value (Vccref), wherein the non-linear mathematical function (5b25) is configured such that a ratio of an output to an input increases with increase in magnitude of input, or the non-linear mathematical function (5b26, 5b27) is configured such that a value of an output with respect to an input increases with increase in magnitude of input and the value of the output is restricted within a determined range, to output the positive-side voltage command value (Vpref2) and the negative-side voltage command value (Vnref2) respectively; and

d) restricting the third voltage command value (Vpccref, Vnccref) with an upper limit value and a lower limit value, the upper limit value being calculated by adding a fixed value to each of the positive-side voltage correction value (Vpcorr) and the negative-side voltage correction value (Vncorr), the lower limit value being calculated by subtracting the fixed value from each of the positive-side voltage correction value (Vpcorr) and the negative-side voltage correction value (Vncorr), to output the positive-side voltage command value (Vpref2) and the negative-side voltage command value (Vnref2) respectively.

The power conversion device according to claim 1, option i), wherein: - the energy storage is a capacitor (1e);

- the control device further includes a first capacitor voltage control portion (5q) configured to generate, separately for each leg circuit, a first voltage correction value (Vpcorr2) based on a difference between a representative value (Vcpu, Vcpv, Vcpw) of a capacitor voltage of the plurality of converter cells constituting the first arm (13) and a command value of the capacitor voltage through feedback control to decrease the difference;

- the circulating current control portion (5b) being configured to generate, separately for each leg circuit, the positive-side voltage command value (Vpref2) with the fourth voltage command value (Vpref, Vpcorr) corrected by linear combination with the first voltage correction value (Vpcorr2);

- the control device further includes a second capacitor voltage control portion (5r) configured to generate, separately for each leg circuit, a second voltage correction value (Vncorr2) based on a difference between a representative value (Vcnu, Vcnv, Vcnw) of a capacitor voltage of the plurality of converter cells constituting the second arm (14) and a command value of the capacitor voltage through feedback control to decrease the difference; and

- the circulating current control portion (5b) being configured to generate, separately for each leg circuit, the negative-side voltage command value (Vnref2) using the fifth voltage command value (Vnref, Vncorr) corrected by linear combination with the second voltage correction value (Vncorr2).

The power conversion device according to claim 2, wherein: - the first capacitor voltage control portion (5q) is configured to generate, separately for each leg circuit, the first voltage correction value (Vpcorr2) by multiplying the difference between the representative value (Vcpu, Vcpv, Vcpw) of the capacitor voltage of the plurality of converter cells constituting the first arm (13) and the command value of the capacitor voltage by a DC current value (Idc) of the DC circuit or a polarity of the DC current value; and

- the second capacitor voltage control portion (5r) is configured to generate, separately for each leg circuit, the second voltage correction value (Vncorr2) by multiplying the difference between the representative value (Vcnu, Vcnv, Vcnw) of the capacitor voltage of the plurality of converter cells constituting the second arm (14) and the command value of the capacitor voltage by the DC current value (Idc) of the DC circuit or the polarity of the DC current value (Idc).

The power conversion device according to claim 2, wherein: - the first capacitor voltage control portion (5q) is configured to generate, separately for each leg circuit, the first voltage correction value (Vpcorr2) by multiplying the difference between the representative value (Vcpu, Vcpv, Vcpw) of the capacitor voltage of the plurality of converter cells constituting the first arm (13) and the command value of the capacitor voltage by an AC current value (Iac) of the AC circuit or a polarity of the AC current value (Iac); and

- the second capacitor voltage control portion (5r) is configured to generate, separately for each leg circuit, the second voltage correction value (Vncorr2) by multiplying the difference between the representative value (Vcnu, Vcnv, Vcnw) of the capacitor voltage of the plurality of converter cells constituting the second arm (14) and the command value of the capacitor voltage by an AC current value obtained by inverting a polarity of the AC current value (Iac) of the AC circuit or a polarity reverse to the polarity of the AC current value (Iac).

Leistungsumwandlungsvorrichtung, die dazu konfiguriert ist, Leistung zwischen einem Gleichstromkreis (4) und einem Wechselstromkreis (2) umzuwandeln, wobei die Leistungsumwandlungsvorrichtung Folgendes umfasst: - eine Vielzahl von Zweigschaltungen (8a, 8b, 8c), die jeweiligen Phasen des Wechselstromkreises (2) entsprechen und zwischen gemeinsamen ersten und zweiten Gleichstromanschlüssen (Np, Nn) parallel geschaltet sind, wobei jede Zweigschaltung in einen hochpotentialseitigen ersten Arm (13) und einen niederpotentialseitigen zweiten Arm (14) mit einem dazwischenliegenden Verbindungsabschnitt (Nu, Nv, Nw), der mit einer entsprechenden Phase des Wechselstromkreises elektrisch verbindbar ist, geteilt ist;
wobei jeder von dem ersten Arm (13) und dem zweiten Arm (14) jeweils Folgendes beinhaltet: - eine Vielzahl von Wandlerzellen (1), die zueinander kaskadiert sind und jeweils einen Energiespeicher (1e) beinhalten, wobei die Vielzahl von Wandlerzellen (1) eine Vielzahl von ersten Wandlerzellen (6a, 6b) und eine Vielzahl von zweiten Wandlerzellen (6c, 6d) beinhaltet; und

- mindestens eine zu der Vielzahl von Wandlerzellen in Reihe geschaltete Induktivität (7a, 7b); wobei die Leistungsumwandlungsvorrichtung Folgendes beinhaltet: - eine Steuerungsvorrichtung (5), die dazu konfiguriert ist, Arbeitsprozesse der Vielzahl von Wandlerzellen zu steuern, wobei die Steuerungsvorrichtung Folgendes beinhaltet: - einen Wechselstromsteuerungsteil (5c), der dazu konfiguriert ist, für jede Zweigschaltung separat einen ersten Spannungsbefehlswert (Vacref), der eine von der Vielzahl von ersten Wandlerzellen (6a, 6b) der Zweigschaltung basierend auf einem Wechselstrom (Iac) und einer Wechselspannung (Vac) des Wechselstromkreises auszugebende Wechselspannungskomponente repräsentiert, zu erzeugen;

- einen Gleichstromsteuerungsteil (5d), der dazu konfiguriert ist, einen zweiten Spannungsbefehlswert (Vdcref), der eine von der Vielzahl von ersten Wandlerzellen (6a, 6b) jeder Zweigschaltung basierend auf einem Gleichstrom (Idc) und einer Gleichspannung (Vdc) des Gleichstromkreises auszugebende Gleichspannungskomponente repräsentiert, zu erzeugen; und

- einen Kreisstromsteuerungsteil (5b), der dazu konfiguriert ist, für jede Zweigschaltung separat einen dritten Spannungsbefehlswert (Vccref), der eine von der Vielzahl von zweiten Wandlerzellen (6c, 6d) auszugebende Spannung repräsentiert, zu erzeugen, um einen Kreisstrom (Icc) für die Zweigschaltung basierend auf dem durch die Zweigschaltung fließenden Kreisstrom (Icc) zu unterdrücken,
wobei i) die Steuerungsvorrichtung (5) einen ersten Befehlswertkombinationsteil (5e) beinhaltet, wobei der erste Befehlswertkombinationsteil (5e) dazu konfiguriert ist, für jede Zweigschaltung separat basierend auf der Differenz zwischen dem ersten Spannungsbefehlswert (Vacref) und dem zweiten Spannungsbefehlswert (Vdcref) einen vierten Spannungsbefehlswert (Vpref) zu berechnen; und die Steuerungsvorrichtung (5) einen zweiten Befehlswertkombinationsteil (5f) beinhaltet, wobei der zweite Befehlswertkombinationsteil (5f) dazu konfiguriert ist, für jede Zweigschaltung separat basierend auf der Summe des ersten Spannungsbefehlswerts (Vacref) und des zweiten Spannungsbefehlswerts (Vdcref) einen fünften Spannungsbefehlswert (Vnref) zu berechnen;

und wobei eine Verstärkungsschaltung (5g) dazu konfiguriert ist, für jede Zweigschaltung separat den vierten Spannungsbefehlswert (Vpref) mit einer Verstärkung (K) zu multiplizieren und einen positivseitigen Spannungskorrekturwert (Vpcorr) an den Kreisstromsteuerungsteil (5b) auszugeben, und eine Verstärkungsschaltung (5h) dazu konfiguriert ist, für jede Zweigschaltung separat den fünften Spannungsbefehlswert (Vnref) mit einer Verstärkung (K) zu multiplizieren und einen negativseitigen Spannungskorrekturwert (Vncorr) an den Kreisstromsteuerungsteil (5b) auszugeben, wobei der Kreisstromsteuerungsteil (5b) dazu konfiguriert ist, für jede Zweigschaltung separat einen positivseitigen Spannungsbefehlswert (Vpref2) zum Betreiben der Vielzahl von zweiten Wandlerzellen (6c) des ersten Arms (13) und einen negativseitigen Spannungsbefehlswert (Vnref2) zum Betreiben der Vielzahl von zweiten Wandlerzellen (6d) des zweiten Arms (14) zu erzeugen und auszugeben, indem er an dem positivseitigen Spannungskorrekturwert (Vpcorr), dem negativseitigen Spannungskorrekturwert (Vncorr) und dem dritten Spannungsbefehlswert (Vccref, Vpccref, Vnccref) eine spezifische Operation durchführt; oder

ii) wobei die Steuerungsvorrichtung (5) einen Kondensatorspannungssteuerungsteil (5q, 5r) beinhaltet, der dazu konfiguriert ist, für jede Zweigschaltung separat einen ersten Spannungskorrekturwert (Vpcorr2) und einen zweiten Spannungskorrekturwert (Vncorr2) basierend auf einem Wert zu erzeugen, der durch Multiplizieren einer Differenz zwischen einem repräsentativen Wert (Vcpu, Vcpv, Vcpw, Vcnu, Vcnv, Vcnw) einer Kondensatorspannung der Vielzahl von zweiten Wandlerzellen (6c, 6d) und einem Befehlswert der Kondensatorspannung mit einem Wechselstromwert des Wechselstromkreises durch Rückkopplungsregelung berechnet wird, um die Differenz zu verringern; und wobei ein erster Addierer (5i) dazu konfiguriert ist, für jede Zweigschaltung separat den zweiten Spannungsbefehlswert (Vdcref) zu dem ersten Spannungskorrekturwert (Vpcorr2) zu addieren und für jede Zweigschaltung einen positivseitigen Spannungskorrekturwert (Vpcorr) an den Kreisstromsteuerungsteil (5b) auszugeben, und ein zweiter Addierer (5j) dazu konfiguriert ist, für jede Zweigschaltung separat den zweiten Spannungsbefehlswert (Vdcref) zu dem zweiten Spannungskorrekturwert (Vncorr2) zu addieren und für jede Zweigschaltung einen negativseitigen Spannungskorrekturwert (Vncorr) an den Kreisstromsteuerungsteil (5b) auszugeben, wobei der Kreisstromsteuerungsteil (5b) dazu konfiguriert ist, für jede Zweigschaltung separat einen positivseitigen Spannungsbefehlswert (Vpref2) zum Betreiben der Vielzahl von zweiten Wandlerzellen (6c) des ersten Arms (13) und einen negativseitigen Spannungsbefehlswert (Vnref2) zum Betreiben der Vielzahl von zweiten Wandlerzellen (6d) des zweiten Arms (14) zu erzeugen und auszugeben, indem er an dem positivseitigen Spannungskorrekturwert (Vpcorr), dem negativseitigen Spannungskorrekturwert (Vncorr) und dem dritten Spannungsbefehlswert (Vccref) eine spezifische Operation durchführt;

- wobei die spezifische Operation gemäß einer der obigen Optionen i) oder ii) eines der Folgenden beinhaltet: a) Multiplizieren des positivseitigen Spannungskorrekturwerts (Vpcorr) und des negativseitigen Spannungskorrekturwerts (Vncorr) mit einer Summe aus dem dritten Spannungsbefehlswert (Vccref) und einem festen Wert, um den positivseitigen Spannungsbefehlswert (Vpref2) bzw. den negativseitigen Spannungsbefehlswert (Vnref2) auszugeben;

b) Addieren eines durch Anwenden einer nichtlinearen Übertragungsfunktion (5b22) auf den dritten Spannungsbefehlswert (Vccref) berechneten Wertes, wobei die nichtlineare Übertragungsfunktion (5b22) derart konfiguriert ist, dass ein Verhältnis eines Ausgangs zu einem Eingang ansteigt, wenn der Eingang einen Schwellenwert überschreitet, zu jedem von dem positivseitigen Spannungskorrekturwert (Vpcorr) und dem negativseitigen Spannungskorrekturwert (Vncorr), um den positivseitigen Spannungsbefehlswert (Vpref2) bzw. den negativseitigen Spannungsbefehlswert (Vnref2) auszugeben;

c) Addieren eines durch Anwenden einer nichtlinearen mathematischen Funktion (5b25) auf den dritten Spannungsbefehlswert (Vccref) erhaltenen Wertes, wobei die nichtlineare mathematische Funktion (5b25) derart konfiguriert ist, dass ein Verhältnis eines Ausgangs zu einem Eingang mit einem Anstieg der Größenordnung des Eingangs ansteigt, oder die nichtlineare mathematische Funktion (5b26, 5b27) derart konfiguriert ist, dass ein Wert eines Ausgangs in Bezug auf einen Eingang mit einem Anstieg der Größenordnung des Eingangs ansteigt und der Wert des Ausgangs innerhalb eines bestimmten Bereichs begrenzt ist, zu jedem von dem positivseitigen Spannungskorrekturwert (Vpcorr) und dem negativseitigen Spannungskorrekturwert (Vncorr), um den positivseitigen Spannungsbefehlswert (Vpref2) bzw. den negativseitigen Spannungsbefehlswert (Vnref2) auszugeben; und

d) Begrenzen des dritten Spannungsbefehlswerts (Vpccref, Vnccref) durch einen oberen Grenzwert und einen unteren Grenzwert, wobei der obere Grenzwert durch Addieren eines festen Werts zu jedem von dem positivseitigen Spannungskorrekturwert (Vpcorr) und dem negativseitigen Spannungskorrekturwert (Vncorr) berechnet wird, wobei der untere Grenzwert durch Subtrahieren des festen Wertes von jedem von dem positivseitigen Spannungskorrekturwert (Vpcorr) und dem negativseitigen Spannungskorrekturwert (Vncorr) berechnet wird, um den positivseitigen Spannungsbefehlswert (Vpref2) bzw. den negativseitigen Spannungsbefehlswert (Vnref2) auszugeben.

Die Leistungsumwandlungsvorrichtung nach Anspruch 1, Option i), wobei: - der Energiespeicher ein Kondensator (1e) ist;

- die Steuerungsvorrichtung ferner einen ersten Kondensatorspannungssteuerungsteil (5q) beinhaltet, der dazu konfiguriert ist, für jede Zweigschaltung separat einen ersten Spannungskorrekturwert (Vpcorr2) basierend auf einer Differenz zwischen einem repräsentativen Wert (Vcpu, Vcpv, Vcpw) einer Kondensatorspannung der Vielzahl von Wandlerzellen, aus denen der erste Arm (13) besteht, und einem Befehlswert der Kondensatorspannung durch Rückkopplungsregelung zu berechnen, um die Differenz zu verringern;

- der Kreisstromsteuerungsteil (5b) dazu konfiguriert ist, für jede Zweigschaltung separat den positivseitigen Spannungsbefehlswert (Vpref2) zu erzeugen, wobei der vierte Spannungsbefehlswert (Vpref, Vpcorr) durch lineare Kombination mit dem ersten Spannungskorrekturwert (Vpcorr2) korrigiert wird;

- die Steuerungsvorrichtung ferner einen zweiten Kondensatorspannungssteuerungsteil (5r) beinhaltet, der dazu konfiguriert ist, für jede Zweigschaltung separat einen zweiten Spannungskorrekturwert (Vncorr2) basierend auf einer Differenz zwischen einem repräsentativen Wert (Vcnu, Vcnv, Vcnw) einer Kondensatorspannung der Vielzahl von Wandlerzellen, aus denen der zweite Arm (14) besteht, und einem Befehlswert der Kondensatorspannung durch Rückkopplungsregelung zu berechnen, um die Differenz zu verringern; und

- der Kreisstromsteuerungsteil (5b) dazu konfiguriert ist, für jede Zweigschaltung separat den negativseitigen Spannungsbefehlswert (Vnref2) unter Verwendung des fünften Spannungsbefehlswerts (Vnref, Vncorr), der durch lineare Kombination mit dem zweiten Spannungskorrekturwert (Vncorr2) korrigiert wurde, zu erzeugen.

Leistungsumwandlungsvorrichtung nach Anspruch 2, wobei: - der erste Kondensatorspannungssteuerungsteil (5q) dazu konfiguriert ist, für jede Zweigschaltung separat den ersten Spannungskorrekturwert (Vpcorr2) durch Multiplizieren der Differenz zwischen dem repräsentativen Wert (Vcpu, Vcpv, Vcpw) der Kondensatorspannung der Vielzahl von Wandlerzellen, aus denen der erste Arm (13) besteht, und dem Befehlswert der Kondensatorspannung mit einem Gleichstromwert (Idc) des Gleichstromkreises oder einer Polarität des Gleichstromwertes zu erzeugen; and

- der zweite Kondensatorspannungssteuerungsteil (5r) dazu konfiguriert ist, für jede Zweigschaltung separat den zweiten Spannungskorrekturwert (Vncorr2) durch Multiplizieren der Differenz zwischen dem repräsentativen Wert (Vcnu, Vcnv, Vcnw) der Kondensatorspannung der Vielzahl von Wandlerzellen, aus denen der zweite Arm (14) besteht, und dem Befehlswert der Kondensatorspannung mit dem Gleichstromwert (Idc) des Gleichstromkreises oder der Polarität des Gleichstromwertes (Idc) zu erzeugen.

Leistungsumwandlungsvorrichtung nach Anspruch 2, wobei: - der erste Kondensatorspannungssteuerungsteil (5q) dazu konfiguriert ist, für jede Zweigschaltung separat den ersten Spannungskorrekturwert (Vpcorr2) durch Multiplizieren der Differenz zwischen dem repräsentativen Wert (Vcpu, Vcpv, Vcpw) der Kondensatorspannung der Vielzahl von Wandlerzellen, aus denen der erste Arm (13) besteht, und dem Befehlswert der Kondensatorspannung mit einem Wechselstromwert (Iac) des Wechselstromkreises oder einer Polarität des Wechselstromwertes (Iac) zu erzeugen; and

- der zweite Kondensatorspannungssteuerungsteil (5r) dazu konfiguriert ist, für jede Zweigschaltung separat den zweiten Spannungskorrekturwert (Vncorr2) durch Multiplizieren der Differenz zwischen dem repräsentativen Wert (Vcnu, Vcnv, Vcnw) der Kondensatorspannung der Vielzahl von Wandlerzellen, aus denen der zweite Arm (14) besteht, und dem Befehlswert der Kondensatorspannung mit einem durch Umkehren einer Polarität des Wechselstromwerts (Iac) des Wechselstromkreises erhaltenen Wechselstromwert oder einer zur Polarität des Wechselstromwertes (Iac) umgekehrten Polarität zu erzeugen.

Dispositif de conversion de puissance configuré pour convertir une puissance entre un circuit CC (4) et un circuit CA (2), le dispositif de conversion de puissance comprenant : - une pluralité de circuits de branche (8a, 8b, 8c) qui correspondent à des phases respectives du circuit CA (2) et sont connectés en parallèle entre des première et seconde bornes CC communes (Np, Nn), dans lequel chaque circuit de branche est divisé en un premier bras (13) du côté haut potentiel et en un second bras (14) du côté bas potentiel avec une portion de connexion interposée (Nu, Nv, Nw) qui peut être électriquement connectée à une phase correspondante du circuit CA ;
chacun du premier bras (13) et du second bras (14) incluant : - une pluralité de cellules de convertisseur (1) montées en cascade les unes par rapport aux autres et incluant chacune un stockage d'énergie (1e), la pluralité de cellules de convertisseur (1) incluant une pluralité de premières cellules de convertisseur (6a, 6b) et une pluralité de secondes cellules de convertisseur (6c, 6d) ; et

- au moins une inductance (7a, 7b) connectée en série à la pluralité de cellules de convertisseur ; le dispositif de conversion de puissance incluant : - un dispositif de régulation (5) configuré pour réguler des opérations de la pluralité de cellules de convertisseur, le dispositif de régulation incluant : - une portion de régulation CA (5c) configurée pour générer, séparément pour chaque circuit de branche, une première valeur de commande de tension (Vacref) représentant une composante de tension alternative à délivrer à partir de la pluralité de premières cellules de convertisseur (6a, 6b) du circuit de branche d'après un courant alternatif (Iac) et une tension alternative (Vac) du circuit CA ;

- une portion de régulation CC (5d) configurée pour générer une deuxième valeur de commande de tension (Vdcref) représentant une composante de tension continue à délivrer à partir de la pluralité de premières cellules de convertisseur (6a, 6b) de chaque circuit de branche d'après un courant continu (Idc) et une tension continue (Vdc) du circuit CC ; et

- une portion de régulation de courant de circulation (5b) configurée pour générer, séparément pour chaque circuit de branche, une troisième valeur de commande de tension (Vccref) représentant une tension à délivrer à partir de la pluralité de secondes cellules de convertisseur (6c, 6d) afin de supprimer un courant de circulation (Icc) pour le circuit de branche d'après le courant de circulation (Icc) qui circule à travers le circuit de branche,
dans lequel i) le dispositif de régulation (5) inclut une première portion de combinaison de valeurs de commande (5e), dans lequel la première portion de combinaison de valeurs de commande (5e) est configurée pour calculer, séparément pour chaque circuit de branche, une quatrième valeur de commande de tension (Vpref) d'après la différence entre la première valeur de commande de tension (Vacref) et la deuxième valeur de commande de tension (Vdcref) ; et le dispositif de régulation (5) inclut une seconde portion de combinaison de valeurs de commande (5f), dans lequel la seconde portion de combinaison de valeurs de commande (5f) est configurée pour calculer, séparément pour chaque circuit de branche, une cinquième valeur de commande de tension (Vnref) d'après la somme de la première valeur de commande de tension (Vacref) et de la deuxième valeur de commande de tension (Vdcref) ;

et dans lequel un circuit de gain (5g) est configuré pour multiplier, séparément pour chaque circuit de branche, la quatrième valeur de commande de tension (Vpref) par un gain (K) et pour délivrer une valeur de correction de tension côté positif (Vpcorr) à la portion de régulation de courant de circulation (5b), et un circuit de gain (5h) est configuré pour multiplier, séparément pour chaque circuit de branche, la cinquième valeur de commande de tension (Vnref) par un gain (K) et pour délivrer une valeur de correction de tension côté négatif (Vncorr) à la portion de régulation de courant de circulation (5b), dans lequel la portion de régulation de courant de circulation (5b) est configurée pour générer et délivrer, séparément pour chaque circuit de branche, une valeur de commande de tension côté positif (Vpref2) pour faire fonctionner la pluralité de secondes cellules de convertisseur (6c) du premier bras (13) et une valeur de commande de tension côté négatif (Vnref2) pour faire fonctionner la pluralité de secondes cellules de convertisseur (6d) du second bras (14) en effectuant une opération spécifique avec la valeur de correction de tension côté positif (Vpcorr), la valeur de correction de tension côté négatif (Vncorr), et la troisième valeur de commande de tension (Vccref, Vpccref, Vnccref) ; ou

ii) dans lequel le dispositif de régulation (5) inclut une portion de régulation de tension de condensateur (5q, 5r) qui est configurée pour générer, séparément pour chaque circuit de branche, une première valeur de correction de tension (Vpcorr2) et une seconde valeur de correction de tension (Vncorr2) d'après une valeur calculée en multipliant une différence entre une valeur représentative (Vcpu, Vcpv, Vcpw, Venu, Vcnv, Vcnw) d'une tension de condensateur de la pluralité de secondes cellules de convertisseur (6c, 6d) et une valeur de commande de la tension de condensateur par une valeur de courant alternatif du circuit CA par le biais d'un asservissement pour diminuer la différence ; et un premier additionneur (5i) est configuré pour ajouter, séparément pour chaque circuit de branche, la deuxième valeur de commande de tension (Vdcref) à ladite première valeur de correction de tension (Vpcorr2) et pour délivrer, pour chaque circuit de branche, une valeur de correction de tension côté positif (Vpcorr) à la portion de régulation de courant de circulation (5b), et un second additionneur (5j) est configuré pour ajouter, séparément pour chaque circuit de branche, la deuxième valeur de commande de tension (Vdcref) à ladite seconde valeur de correction de tension (Vncorr2) et pour délivrer, pour chaque circuit de branche, une valeur de correction de tension côté négatif (Vncorr) à la portion de régulation de courant de circulation (5b), dans lequel la portion de régulation de courant de circulation (5b) est configurée pour générer et délivrer, séparément pour chaque circuit de branche, une valeur de commande de tension côté positif (Vpref2) pour faire fonctionner la pluralité de secondes cellules de convertisseur (6c) du premier bras (13) et une valeur de commande de tension côté négatif (Vnref2) pour faire fonctionner la pluralité de secondes cellules de convertisseur (6d) du second bras (14) en effectuant une opération spécifique avec la valeur de correction de tension côté positif (Vpcorr), la valeur de correction de tension côté négatif (Vncorr) et la troisième valeur de commande de tension (Vccref) ;

- l'opération spécifique selon l'une des options i) ou ii) ci-dessus incluant l'un quelconque parmi : a) la multiplication de chacune de la valeur de correction de tension côté positif (Vpcorr) et de la valeur de correction de tension côté négatif (Vncorr) par une somme de la troisième valeur de commande de tension (Vccref) et d'une valeur fixe, pour délivrer la valeur de commande de tension côté positif (Vpref2) et la valeur de commande de tension côté négatif (Vnref2) respectivement ;

b) l'ajout, à chacune de la valeur de correction de tension côté positif (Vpcorr) et de la valeur de correction de tension côté négatif (Vncorr), d'une valeur calculée en appliquant une fonction de transfert non linéaire (5b22) à la troisième valeur de commande de tension (Vccref), dans lequel la fonction de transfert non linéaire (5b22) est configurée de telle sorte qu'un rapport entre une sortie et une entrée augmente lorsque l'entrée dépasse une valeur seuil, pour délivrer la valeur de commande de tension côté positif (Vpref2) et la valeur de commande de tension côté négatif (Vnref2), respectivement ;

c) l'ajout, à chacune de la valeur de correction de tension côté positif (Vpcorr) et de la valeur de correction de tension côté négatif (Vncorr), d'une valeur obtenue en appliquant une fonction mathématique non linéaire (5b25) à la troisième valeur de commande de tension (Vccref), dans lequel la fonction mathématique non linéaire (5b25) est configurée de telle sorte qu'un rapport entre une sortie et une entrée augmente avec l'augmentation de la grandeur de l'entrée, ou la fonction mathématique non linéaire (5b26, 5b27) est configurée de telle sorte qu'une valeur d'une sortie par rapport à une entrée augmente avec l'augmentation de la grandeur de l'entrée et que la valeur de la sortie est restreinte dans une plage déterminée, pour délivrer la valeur de commande de tension côté positif (Vpref2) et la valeur de commande de tension côté négatif (Vnref2) respectivement ; et

d) la restriction de la troisième valeur de commande de tension (Vpccref, Vnccref) avec une valeur limite supérieure et une valeur limite inférieure, la valeur limite supérieure étant calculée en ajoutant une valeur fixe à chacune de la valeur de correction de tension côté positif (Vpcorr) et de la valeur de correction de tension côté négatif (Vncorr), la valeur limite inférieure étant calculée en soustrayant la valeur fixe de chacune de la valeur de correction de tension côté positif (Vpcorr) et de la valeur de correction de tension côté négatif (Vncorr), pour délivrer la valeur de commande de tension côté positif (Vpref2) et la valeur de commande de tension côté négatif (Vnref2) respectivement.

Dispositif de conversion de puissance selon la revendication 1, option i), dans lequel : - le stockage d'énergie est un condensateur (1e) ;

- le dispositif de régulation inclut en outre une première portion de régulation de tension de condensateur (5q) configurée pour générer, séparément pour chaque circuit de branche, une première valeur de correction de tension (Vpcorr2) d'après une différence entre une valeur représentative (Vcpu, Vcpv, Vcpw) d'une tension de condensateur de la pluralité de cellules de convertisseur constituant le premier bras (13) et une valeur de commande de la tension de condensateur par le biais d'un asservissement pour diminuer la différence ;

- la portion de régulation de courant de circulation (5b) étant configurée pour générer, séparément pour chaque circuit de branche, la valeur de commande de tension côté positif (Vpref2) avec la quatrième valeur de commande de tension (Vpref, Vpcorr) corrigée par combinaison linéaire avec la première valeur de correction de tension (Vpcorr2) ;

- le dispositif de régulation inclut en outre une seconde portion de régulation de tension de condensateur (5r) configurée pour générer, séparément pour chaque circuit de branche, une seconde valeur de correction de tension (Vncorr2) d'après une différence entre une valeur représentative (Venu, Vcnv, Vcnw) d'une tension de condensateur de la pluralité de cellules de convertisseur constituant le second bras (14) et une valeur de commande de la tension de condensateur par le biais d'un asservissement pour diminuer la différence ; et

- la portion de régulation de courant de circulation (5b) étant configurée pour générer, séparément pour chaque circuit de branche, la valeur de commande de tension côté négatif (Vnref2) en utilisant la cinquième valeur de commande de tension (Vnref, Vncorr) corrigée par combinaison linéaire avec la seconde valeur de correction de tension (Vncorr2).

Dispositif de conversion de puissance selon la revendication 2, dans lequel : - la première portion de régulation de tension de condensateur (5q) est configurée pour générer, séparément pour chaque circuit de branche, la première valeur de correction de tension (Vpcorr2) en multipliant la différence entre la valeur représentative (Vcpu, Vcpv, Vcpw) de la tension de condensateur de la pluralité de cellules de convertisseur constituant le premier bras (13) et la valeur de commande de la tension de condensateur par une valeur de courant continu (Idc) du circuit CC ou une polarité de la valeur de courant continu ; et

- la seconde portion de régulation de tension de condensateur (5r) est configurée pour générer, séparément pour chaque circuit de branche, la seconde valeur de correction de tension (Vncorr2) en multipliant la différence entre la valeur représentative (Venu, Vcnv, Vcnw) de la tension de condensateur de la pluralité de cellules de convertisseur constituant le second bras (14) et la valeur de commande de la tension de condensateur par la valeur de courant continu (Idc) du circuit CC ou la polarité de la valeur de courant continu (Idc).

Dispositif de conversion de puissance selon la revendication 2, dans lequel : - la première portion de régulation de tension de condensateur (5q) est configurée pour générer, séparément pour chaque circuit de branche, la première valeur de correction de tension (Vpcorr2) en multipliant la différence entre la valeur représentative (Vcpu, Vcpv, Vcpw) de la tension de condensateur de la pluralité de cellules de convertisseur constituant le premier bras (13) et la valeur de commande de la tension de condensateur par une valeur de courant alternatif (Iac) du circuit CA ou une polarité de la valeur de courant alternatif (Iac) ; et

- la seconde portion de régulation de tension de condensateur (5r) est configurée pour générer, séparément pour chaque circuit de branche, la seconde valeur de correction de tension (Vncorr2) en multipliant la différence entre la valeur représentative (Venu, Vcnv, Vcnw) de la tension de condensateur de la pluralité de cellules de convertisseur constituant le second bras (14) et la valeur de commande de la tension de condensateur par une valeur de courant alternatif obtenue en inversant une polarité de la valeur de courant alternatif (Iac) du circuit CA ou une polarité inverse à la polarité de la valeur de courant alternatif (Iac).

REFERENCES CITED IN THE DESCRIPTION

This list of references cited by the applicant is for the reader's convenience only. It does not form part of the European patent document. Even though great care has been taken in compiling the references, errors or omissions cannot be excluded and the EPO disclaims all liability in this regard.

Patent documents cited in the description

JP5189105B[0004][0005][0011][0012]
JP2012531878A[0004][0006][0011][0014]
WO2011042050A1[0010]

Non-patent literature cited in the description

A review an circulating current suppression control, capacitor voltage balancing and fault analysis of modular multilevel convertersSAI PRASANNA P et al.2015 INTERNATIONAL CONFERENCE ON ELECTRICAL, ELECTRONICS, SIGNALS, COMMUNICATION AND OPTIMIZATION (EESCO)IEEE2015012416
[0008]