(19)
(11)EP 2 649 717 B1

(12)EUROPEAN PATENT SPECIFICATION

(45)Mention of the grant of the patent:
09.09.2020 Bulletin 2020/37

(21)Application number: 11848715.6

(22)Date of filing:  12.12.2011
(51)International Patent Classification (IPC): 
G01R 31/40(2020.01)
H02M 7/48(2007.01)
G01R 31/42(2006.01)
(86)International application number:
PCT/US2011/064476
(87)International publication number:
WO 2012/082648 (21.06.2012 Gazette  2012/25)

(54)

METHOD AND SYSTEM FOR MEASURING THE INTEGRITY OF A POWER CONVERTER

VERFAHREN UND SYSTEM ZUR MESSUNG DER INTAKTHEIT EINES STROMRICHTERS

PROCÉDÉ ET SYSTÈME POUR MESURER L'INTÉGRITÉ D'UN CONVERTISSEUR DE PUISSANCE


(84)Designated Contracting States:
AL AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HR HU IE IS IT LI LT LU LV MC MK MT NL NO PL PT RO RS SE SI SK SM TR

(30)Priority: 12.12.2010 US 422194 P

(43)Date of publication of application:
16.10.2013 Bulletin 2013/42

(73)Proprietor: Infinirel Corporation
Plano, TX 75025 (US)

(72)Inventors:
  • WANK, Norbert, H.
    Plano, TX 75025 (US)
  • FRIEDMAN, David H.
    Plano, TX 75025 (US)

(74)Representative: Bennett, Adrian Robert J. et al
A.A. Thornton & Co. 15 Old Bailey
London EC4M 7EF
London EC4M 7EF (GB)


(56)References cited: : 
EP-A1- 1 630 937
WO-A2-2010/111412
JP-A- H08 171 430
US-A1- 2008 285 184
US-B2- 7 512 825
WO-A2-2010/111412
WO-A2-2010/138948
US-A1- 2007 121 795
US-A1- 2010 161 259
  
      
    Note: Within nine months from the publication of the mention of the grant of the European patent, any person may give notice to the European Patent Office of opposition to the European patent granted. Notice of opposition shall be filed in a written reasoned statement. It shall not be deemed to have been filed until the opposition fee has been paid. (Art. 99(1) European Patent Convention).


    Description

    TECHNICAL FIELD



    [0001] The present disclosure relates generally to electronic power converters and, in an embodiment described herein, more particularly provides a system and method for non-intrusively detecting degradation in power output quality from power conversion equipment, thereby predicting impending failures in the power conversion equipment.

    BACKGROUND



    [0002] Failures in power conversion systems, such as power converters failing while producing power to a power grid, can cause emergency situations by removing a power source to the power grid with minimal warnings. Emergencies (e.g. unanticipated maintenance, power loss, etc.) can be very expensive for both power customers and power companies. Minimizing on-grid failures could reduce emergency maintenance expenses, allow more efficient spare parts ordering and storage, reduce power grid outages, etc. Therefore, it can readily be seen that improvements to the art of power conversion systems is needed.

    [0003] Therefore, it can readily be seen that improvements to the art of power conversion systems is needed.

    [0004] EP1630937 discloses a voltage conversion device and judging occurrence of a failure in the device.

    SUMMARY



    [0005] In carrying out the principles of the present disclosure, a method and system is provided which brings improvements to the art of power conversion.

    [0006] The invention provides a method of measuring integrity of a pulse-width modulated power converter according to claim 1 and a system for measuring an integrity of a pulse-width modulated power converter according to claim 13. Preferred embodiments are defined in the dependent claims.

    [0007] One example is described below in which a method of measuring integrity of a pulse-width modulated power converter may include the steps of monitoring data associated with an input and an output of the power converter. Then correlating the data to identify degradation in performance of the power converter and indicate impending failures of the power converter. Another example is described below in which a system for measuring integrity of a pulse-width modulated power converter is provided. The system may include a power converter and current sensors that sense input and output currents of the power converter. A programmable controller may be loaded with an integrity measuring algorithm that may analyze data from the current sensors. An impending failure of the power converter may be indicated as a result of the data analysis.

    [0008] These and other features, advantages and benefits will become apparent to one of ordinary skill in the art upon careful consideration of the detailed description of representative embodiments of the disclosure below and the accompanying drawings, in which similar elements are indicated in the various figures using the same reference numbers.

    BRIEF DESCRIPTION OF THE DRAWINGS



    [0009] 

    FIG. 1 is a representative block diagram of a prior art power conversion system which may benefit from the principles of the present disclosure.

    FIG. 2 is a representative block diagram of a power conversion system which embodies principles of the present disclosure.

    FIG. 3 is a representative block diagram of an algorithm which calculates an equivalent series resistance.

    FIG. 4 is a representative block diagram of another algorithm which calculates an equivalent series resistance.

    FIG. 5 is a schematic representation of electrical components of a power converter.

    FIG. 6 is a schematic representation of output circuitry of the power converter.

    FIGS. 6a-c are waveforms of parameters of the power converter and an integrity measuring algorithm.

    FIG. 7 is a schematic representation of input circuitry of the power converter.

    FIGS. 7a-f are waveforms of additional parameters of the power converter and the integrity measuring algorithm.

    FIG. 8 is a representative flow chart of a method for measuring integrity of a power converter.


    DETAILED DESCRIPTION



    [0010] Generally there are four types of power conversions, as given in the table below.
    TopologyInputOutputTypical Application
    inverter DC AC PV solar, wind turbines, AC-motor (Variable-Speed) drives
    DC-DC converter DC DC telecom and data-center power, servers, battery chargers
    frequency converter AC AC light-rail, heavy rail (locomotive), air-planes, military weapon systems
    rectifier AC DC wind-turbines, electric vehicles (regenerative breaking)


    [0011] A power converter that converts DC power to AC power is normally referred to as an inverter. A DC-DC power converter generally converts from one DC voltage to another DC voltage. A frequency power converter generally converts AC power at one frequency to AC power at another frequency. A power converter that converts AC power to DC power is generally referred to as a rectifier. The principles of this disclosure may be used with any of these four ways of power conversion to indicate impending failures in a power converter.

    [0012] Representatively illustrated in FIG. 1 is a prior art power conversion system 10 which can benefit from the principles of this disclosure. The power conversion system may include a power source 12 that supplies AC and/or DC power to the inputs 48 of a power converter 14. The power source 12 may include multiple sources 12 (FIG. 1) or the power source 12 may be connected in parallel with an input 48 of each power converter 14 such that each power converter receives power from the single power source 12.

    [0013] Each power converter 14 may convert the input power it receives to another power on its outputs 28 by converting the voltage levels, frequencies, etc. and outputting the resulting power to a power grid 30. This power grid 30 may be a national power grid, such as the power grid used in the United States to distribute power to businesses and individual customers. However, the power grid 30 may be a local power grid, such as a local power bus to a Variable Speed Driver for an AC motor in an Electric Vehicle.

    [0014] FIG. 1 shows multiple power converters 14 connected in parallel with the power grid 30. However, it is not necessary that the converters 14 be connected to a single power grid 30. For example, each of the converters 14 may be connected to a separate power grid 30. Additionally, some of the power converters 14 may be connected in parallel and supplying power to one power grid 30 while other power converters 14 are connected individually to separate power grids 30. Therefore, it can readily be seen that any size power grid 30 may be used and that multiple configurations of connecting inputs 48 and/or outputs 28 of the power converters 14 may be employed in keeping with the principles of this disclosure.

    [0015] FIG. 1 shows a representative block diagram of the power converter 14. A controller 36 may be used to regulate the outputs 28 of the power converter 14 by measuring its input/output voltages and input/output currents with appropriate sensors 16, 18, 22, 24. Sensor data is provided to the controller 36 through conventional analog-to-digital (A/D) converters. Current sensors 18 and 24 may provide input and output current measurements, respectively. Voltage sensors 16 and 22 may provide input and output voltage measurements, respectively. The controller 36 may control a power matrix 26 to convert power from the power source 12 to the power grid 30. The controller 36 also may provide communication to an external control system via the MOD control bus 32.

    [0016] The A/D converters may be provided in the power converter 14 along with the controller 36 when purchased from a manufacturer. These sensors and A/D converters usually provide a sampling frequency in a range of approximately 80kHz - 250kHz which provides sufficient data to the controller 46 for adequately controlling the power conversion. However, these sampling frequencies are generally too slow for implementing the principles of this disclosure.

    [0017] Referring now to FIG. 2, an example power conversion system 20 that embodies principles of this disclosure is shown. As in FIG. 1, the power source 12 may be multiple sources, a single source, a single source feeding multiple power converters, etc. Also the power converters 14 may be connected in parallel to the power grid 30, connected to individual power grids 30, etc.

    [0018] Fig. 2 differs from FIG. 1 in that additional voltage sensors 40, 44 and current sensors 42, 46 are placed on the inputs and outputs of the power converter 14. FIG. 2 shows that only one of the power converters is outfitted with additional sensors 40, 42, 44, 46, however, it can readily be seen that any of the power converters 14 may be configured to incorporate the additional sensors. These sensors 40, 42, 44, 46 provide higher sampling frequencies than those of
    FIG. 1, and a programmable controller 38 includes high-speed A/D converters for converting data from the sensors 40, 42, 44, 46 into digital information for use by the programmable controller 38. It may also be readily seen that multiple of these sensors may be used to provide redundancy, increase bandwidth of data collection, etc.

    [0019] For a power conversion system 20 with a power output frequency of 60 Hz, the desired sampling frequencies for the sensors 40, 42, 44, 46 and A/D converters are greater than 250 kHz, and are preferably 1 MHz. It is not a requirement that the sampling frequencies be greater than 250 kHz. The sampling frequencies may be increased or decreased as needed to accommodate power conversion systems 20 with either higher or lower power output frequencies (e.g. 400 Hz systems for aircraft, 50 Hz for non-US power grids, etc.) in keeping with the principles of this disclosure.

    [0020] These sampling frequencies should be a frequency that provides the programmable controller 38 with visibility into signals within the power conversion system 20. The programmable controller 38 can analyze and detect anomalies in these signals, where the anomalies indicate impending failures of the power converter 14.

    [0021] The programmable controller 38 is loaded with an integrity measuring algorithm 50 (FIGS. 3, 4) that reads the sensor data from sensors 40, 42, 44, 46, correlates the data, and produces an equivalent series resistance (ESR) value 72 from the correlated data. The calculated ESR value 72 may be used to identify anomalies which indicate impending failures of the power converter 14. For example, if the calculated ESR value 72 is significantly larger than the ESR value of an input capacitor of a similar power converter 14, then this may indicate that the input capacitor is going to fail soon.

    [0022] Additionally, if the value 72 is recorded as a function of time (e.g. strip chart recording, digital recording, etc.) and oscillations are detected on the recording. These oscillations may indicate a problem with a software control program, an input capacitor, an insulated gate bipolar transistor, an inductor, a MOSFET device, a bridge switch, etc. of the power converter 14.

    [0023] The voltage sensor 40 reads a value of an input voltage Vin on an input 48 of the power converter 14. The current sensor 42 reads a value of an input current Iin being transferred through input 48 into the power converter 14. The voltage sensor 44 reads a value of an output voltage Vout on an output 28 of the power converter 14. The current sensor 46 reads a value of an output current Iout being transferred through output 28 from the power converter 14.

    [0024] Referring now to FIG. 3, a block diagram of the integrity measuring algorithm 50 is shown. In this version of the algorithm 50, the ESR value 72 is calculated from three inputs 52, 56, 58 which are delivered to the programmable controller 38. The sensors 40, 42, 44, 46 are sampled at a sampling frequency (e.g. 1MHz) and the instantaneous value of each sample is provided to programmable controller 38 via the high-speed A/D converters.

    [0025] Each of the resulting signals Iin, Vin, Iout, Vout include an alternating component and a ripple component. The alternating component is the base frequency of the power system 20. For example, the alternating component would be 60 Hz for a 60 Hz power system 20 and it would be 50 Hz for a 50 Hz power system 20. The ripple component is that part of the signal that represents a charging and discharging of a capacitor. The ripple component may be similar to a sawtooth waveform when separated from the alternating component.

    [0026] A detrending function may be used to separate the alternating component from the ripple component. As seen in FIG. 3, a ripple component of the output current Iout is calculated in step 52 of the algorithm 50. The Iout ripple is then output to step 54 where a first derivative (e.g. first difference (x(n+1)-x(n))/(t(n+1)-t(n))) of the Iout ripple signal is calculated. In step 56, a ripple component of the input current Iin is calculated and output to step 62. In step 58, the alternating component of the output current Iout is calculated (i.e. the ripple component is removed) and then the alternating component is output to the step 62.

    [0027] The step 62 multiplies the results of steps 56 and 58 and then outputs the resulting product to a multiplier step 60. Step 64 calculates an RMS value of the output current Iout alternating component result from step 58 and then step 68 calculates a square of the RMS value from step 64. Step 68 squares the RMS value from step 64 and presents the result to step 70.

    [0028] The step 60 multiplies the results of steps 54 and 62 and then outputs the resulting product to step 66 where a mean value is calculated. Step 66 outputs the mean value to step 70 where the mean value is divided by the squared RMS value from step 68. The output of the step 70 is the equivalent series resistance value 72. The calculated ESR value 72 is used to indicate impending failures of the power converter 14.

    [0029] A beneficial aspect of the algorithm 50 in FIG. 3 is that only the input and output currents Iin, Iout need to be sensed to provide indications of impending failures of the power converter 14. The calculated ESR value 72 from the algorithm of FIG. 3 may have a larger error tolerance of other implementations of the algorithm 50, but it does provide indications of impending failures of the power converter 14.

    [0030] Another implementation of the algorithm 50 is shown in FIG. 4. This implementation produces an ESR value 72 with tighter error tolerances and is capable of more accurate results than the implementation of FIG. 3. FIG. 4 is very similar to FIG. 3 except for a few important differences. Generally, the differences are that step 58 has been replaced by step 74, and steps 76 and 78 have been added.

    [0031] In step 74, the alternating component of the output voltage Vout is calculated by removing the ripple component from the output voltage signal (in step 58 of FIG. 3 the alternating component of the output current Iout was calculated). The step 76 calculates an average of a power factor between the output voltage Vout and current Iout. The result of step 76 is input into step 78. Step 78 multiplies the output of step 76 with the output of step 70. The resulting product is the calculated ESR value 72.

    [0032] Other scaling and sign manipulations can be added to the algorithm 50 while continuing to provide an ESR value 72 which provides indications of impending failures of the power converter 14. Therefore, it can readily be seen that several variations of the algorithm 50 are possible in keeping with the principles of the current disclosure.

    [0033] Referring now to FIG. 5, an example of the power converter 14 is disclosed to aid in a more detailed understanding of an implementation of the algorithm 50. However, it must be understood this is merely an example of the power converter 14 and algorithm 50 and in no way limits the principles of this disclosure.

    [0034] FIG. 5 shows an equivalent circuit for the power converter 14 (which in this example is an inverter) plus a power source 12 and an output load Rload. The power source 12 is represented by a DC voltage source Vdc and a DC resistance Rdc. The power source 12 may be connected to input terminals 80 of the power converter 14, with an input voltage Vin and an input current Iin. A DC link capacitor 82 is represented as C1 and R1.

    [0035] A bridge switch 88 is a switch that alternates between a straight-thru connection 84 and a crossed-over connection 86 between the inputs and outputs of the switch 88. The state of the bridge switch 88 is represented by the value of a variable S which represents the instantaneous state of the switch: S = 1 for the straight-thru connection 84 and S = -1 for the crossed-over connection 86. A current Isi is the current passing through an energy storage inductor L which is on an output side of the bridge switch 88. A corresponding current on an input side of the bridge switch 88 is shown as the product of the variable S and the inductor current Isi (=S*Isi). Therefore, the bridge switch 88 input current value alternates between Isi and -Isi depending on the switch state (i.e. the value of S).

    [0036] A capacitor C2 and resistor R2 represent an output capacitor 90, where an output voltage Vout and an output current Iout is output from the power converter 14. A power converter load 94 is represented by a resistive load component Rload, a current source Ifo for forcing an optional inductive load current component, and additional power converters 14. The additional power converters 14 are shown numbered 1, 2, ... Npar, and are paralleled on an output bus 96 connected to the output terminals 92 of the power converter 14. Each additional power converter 14 is represented by a capacitor C3 connected in series with a resistor R3.

    [0037] The current Isi through the inductor L can be determined as a time-integral of a voltage across the inductor L. As seen in FIG. 6, this voltage is the difference between a source voltage Vs (where Vs = S * Vin) and the output voltage Vout. Referring now to FIG. 6a, a short interval of time consisting of a few switching cycles is shown, during which an alternating component (60 Hz in this example) of the output voltage Vout changes relatively slowly and may be taken as essentially constant.

    [0038] FIG. 6a shows the source voltage Vs and the output voltage Vout plotted vs. time. Since the inductor current Isi is proportional to an integral of the difference between Vs and Vout then a switching-ripple component 100 of the inductor current Isi is as shown in FIG. 6b. Assuming very little of a 60 Hz current flows through the output capacitor, a 60 Hz component of the inductor current Isi (not shown) becomes the 60 Hz load current.

    [0039] The ripple component 100 is essentially divided between the output capacitor 90 (C2 and R2) and the output capacitors 98 (C3 and R3) of the additional power converters 14. Therefore, the output current Iout includes a ripple component 102 that is approximately Npar/(Npar+1) times that of the inductor current Isi, as seen in FIG. 6c.

    [0040] Referring now to FIG. 7, a representative schematic of the circuitry on the inputs 80 of the power converter 14 and a switched inductor current (S * Isi) which flows out of the link capacitor 82. FIG. 7a shows the variable S as a function of time. FIG. 7b shows the switched inductor current 104, which is the inductor current Isi multiplied by the state S of the switching configuration (S * Isi). (note: the dotted lines 106 represent the inductor current Isi without the switching). An RMS value of the switched inductor current is accounted for by the input current Iin.

    [0041] If the resistance Rdc of the power source 12, is high relative to an impedance of the link capacitor 82 at a switching frequency (e.g. 60 Hz), then most of the ripple component of the switched inductor current (S * Isi) flows through the link capacitor 82. This gives rise to a capacitive component Vinc (due to capacitor C1) and a resistive component Vinr (due to resistor R1) in a ripple of the input voltage Vin. FIG.7c shows the capacitive component Vinc as trace 108. FIG.7d shows the resistive component Vinr as trace 110. Capacitive and resistive components will also be present in the input current Iin.

    [0042] The ripple component 102 of the output current Iout seen in FIG. 6c is reproduced in FIG. 7e for convenience. Note that a first derivative of the ripple component 102 (or a first difference (x(n+1)-x(n))/(t(n+1)-t(n))delta t in a sampled-digital domain), shown in FIG. 7f as trace 112, has the correct timing to correlate with the resistive component Vinr (i.e. trace 110 in FIG. 7d) of the input voltage Vin across the link capacitor 82. A sample-by-sample product of the two traces 110 and 112, summed and averaged over time, produces a nonzero result proportional to R1.

    [0043] Since the alternating component of the inductor current Isi becomes the alternating component of the output current Iout, a sign of the inductor current Isi changes over the alternating cycle. Therefore it is desirable to compensate for this by including the alternating component of the output current Iout in the sample-by-sample product of the two traces 110 and 112, and dividing by the mean-square of this component in order to scale the result so as to be independent of the magnitude of the output current Iout.

    [0044] Experimental results from a simulation of the algorithm 50 as shown in FIG. 3 was performed. Using a combination of resistive load Rload and forced inductive load current Ifo (e.g. lagging the output voltage Vout by 90 degrees), the simulation produced ESR values 72 that were substantially constant as a load current magnitude was doubled, even though the ESR values 72 did increase as the load became more inductive. This simulation generated ESR values 72 which were accurate enough to correctly predict impending failures of the power converter 14.

    [0045] Multiplying the ESR values with the output power factor, which is computed by correlating the alternating components of the output voltage Vout and current Iout, can achieve improved accuracy by compensating for the increase in ESR values 72 due to changes in inductive loading on the output bus 96. The variations due to inductive loading changed can be reduced to a few percent as the power factor is changed from .7 to 1. Capacitive loading of the power bus may also be compensated accordingly.

    [0046] The circuit of FIG. 5 was simulated by using the nodal equations below:





    [0047] A circuit state was determined at each sample time according to the following procedure:
    1. 1) Determine the current value of S according to the switching frequency with the duty cycle following a 60 Hz sinusoidal function between extremes of e.g. 10% and 90%.
    2. 2) Solve the nodal equations, using the current values of V1, V2 and Isi where V1 and V2 are the capacitive voltages across C1 and C2 respectively.
    3. 3) Update V1, V2 and Isi for the next sample time according to the currents and voltages obtained from step 2, using:







    [0048] Where ts is the time between samples and the primes (i.e. V1', V2', Isi') indicate the updated values). This kind of simple linear update can introduce errors if the sampling rate is too low; however, with a sampling rate of 240k/sec and switching frequency of 20 kHz, doubling the sampling rate was found to have negligible effect on the results.

    [0049] Component values used were:

    Sampling rate: 240k/sec

    Switching rate: 20 kHz

    Vdc = 240 volts

    Rdc = 10 ohms

    C1 = 1000 µF

    R1 = 0.1 ohm

    Lsi = 1 mH

    C2 = 1 µF

    R2 = 1.0 ohm

    Npar = 2



    [0050] The Vdc value was adjusted as needed to produce the desired 120V RMS output. The R1 value is modified during the simulation as a test parameter to see how the out ESR values 72 track the R1 values. Npar of 2 was chosen which equates to 3 inverters in parallel on the output bus 96.

    [0051] The simulation was run for 24,000 samples (0.1 sec = 6 cycles of 60 Hz). Signals Vin, Iin, Vout, Iout were kept and the 60 Hz and switching-ripple components separated by running a "detrend" function program to subtract the best 3rd-degree polynomial fit to blocks of 200 samples (0.833 msec) with 60-sample overlap, to give the ripple components; subtracting the ripple gave the alternating components (60 Hz in this simulation).

    [0052] The numerical computation carried out was:
    1. 1) at each sample time, take the product of
      1. (a) ripple component of Iin
      2. (b) 60 Hz component of Vout
      3. (c) first difference of ripple component of Iout
    2. 2) sum and average the above products
    3. 3) multiply the result by the output power factor
    4. 4) divide the result by the mean-square of the 60 Hz component of Iout


    [0053] An additional step may be performed to compensate for a linear affect on the resulting values due to a deviation of the RMS output voltage from the nominal value (nominal value is 120V in this simulation). Further dividing the result above by a factor equal to "(actual RMS of Vout / nominal RMS of Vout)" will correct for this linear affect.

    [0054] A computation was done for the last 8000 samples (2 cycles at 60 Hz) of the data to reduce the effect of any startup transient. Numerical results obtained were as follows, where:

    Ires = RMS resistive-load 60Hz output current component (amperes)

    Iind = RMS inductive-load 60Hz output current component (amperes)

    Actual RMS of Vout was held to 120 +/- 1 v by adjusting Vdc.

    R1 = 0.002 ohm

    Ires = 6 Iind = 0 result = -.000010 (.000435)

    Ires = 6 Iind = 6 result = .000002 (.000087)

    Ires = 12 Iind = 0 result = .000002 (.000087)

    Ires = 12 Iind = 12 result = .000011 (.000483)

    R1 = 0.1 ohm

    Ires = 6 Iind = 0 result = .002348 (.100964)

    Ires = 6 Iind = 6 result = .002278 (.099093)

    Ires = 12 Iind = 0 result = .002318 (.100833)

    Ires = 12 Iind = 12 result = .002291 (.099659)

    R1 = 0.2 ohm

    Ires = 6 Iind = 0 result = .004717 (.205190)

    Ires = 6 Iind = 6 result = .004551 (.197969)

    Ires = 12 Iind = 0 result = .004744 (.206364)

    Ires = 12 Iind = 12 result = .004585 (.199448)



    [0055] Raw values are given in the third column. These values may be multiplied by a constant factor (43.5 was used in this simulation) to give the values in parentheses, which are estimates of resistance R1 of the link capacitor 82 in ohms.

    [0056] The simulation described above is merely an example and in no way limits the principles of the current disclosure.

    [0057] FIG. 8, a representatively illustrates a method 120 for measuring integrity of a power converter. The method may be used with the power conversion systems 10, 20, or the method may be used with other power conversion systems in keeping with the principles of this disclosure.

    [0058] In step 122, data Vin, Vout, Iin, Iout may be monitored for the inputs and outputs of the power converter 14. This data is provided to step 124 where the data may be correlated to identify degradation in performance of the power converter 14. The correlation may determine/detect an anomaly in the data. In step 126 a decision is made as to whether or not an anomaly is detected. If so, then proceed on to step 128. If not, then proceed back to step 122 and begin process again.

    [0059] In step 128, the type of anomaly may be determined to be a change in the ESR values 72, an oscillation the ESR value, etc. The anomaly may be analyzed in step 128 to identify if the anomaly and determine the characteristics of an impending failure. Once the impending failure is detected/determined, then a maintenance activity 130 may be scheduled (e.g. initiated by sending a command to an operator or to an automated maintenance system) to repair or replace the component that has been determined to be a culprit of the impending failure. Step 132 represents the successful completion of the maintenance activity, thus averting the potential failure of the power converter 14 while the converter is outputting power to the power grid 30.


    Claims

    1. A method of measuring integrity of a pulse-width modulated power converter (14), the method comprising the steps of:

    monitoring input data associated with an input to the pulse-width modulated power converter (14);

    monitoring output data associated with an output from the pulse-width modulated power converter (14); and

    correlating the input data and the output data to identify a degradation in performance of the power converter in a way that either:

    a) a ripple component of the input current (Iin), a ripple component of the output current (Iout) and an alternating component of the output current (Iout); or

    b) a ripple component of the input current (Iin), a ripple component of the output current (Iout), an alternating component of the output voltage (Vout) and the power factor average;

    are processed to calculate an equivalent series resistance of the power converter.


     
    2. The method of claim 1, wherein the correlating step includes determining a degradation in performance in the correlated input and output data, thereby indicating an impending failure of the power converter (14).
     
    3. The method of claim 2, further comprising step of initiating an automated maintenance activity as a result of the impending failure indication, thereby averting a failure of the power converter while the power converter (14) is outputting power to a power grid (30).
     
    4. The method of claim 2, wherein an anomaly is a substantial change in an equivalent series resistance of the power converter, thereby indicating the impending failure of a component of the power converter.
     
    5. The method of claim 1, wherein the correlating step is performed while the power converter (14) is operational, and wherein the correlating step includes indicating impending failures of the power converter (14), and calculating an equivalent series resistance for the power converter (14).
     
    6. The method of claim 5, wherein the correlating step includes recording the equivalent series resistance as a function of time.
     
    7. The method of claim 5, wherein calculating the equivalent series resistance includes compensating for at least one of temperature changes, power factor changes, and output voltage changes.
     
    8. The method of claim 5, wherein calculating the equivalent series resistance includes averaging a sample-by-sample product of, a ripple of an input current, a first derivative of a ripple of an output current, and an alternating component of one of the output current and an output voltage.
     
    9. The method of claim 1, wherein the monitoring steps include monitoring an input current and an output current, and wherein the correlating step includes calculating an equivalent series resistance from values of the input and output currents.
     
    10. The method of claim 1, wherein the monitoring steps include monitoring an input current, an output current, and an output voltage, and wherein the correlating step includes calculating an equivalent series resistance from values of the input current, the output current, and the output voltage.
     
    11. The method of claim 1, the method further comprising a step of connecting only the output of the pulse-width modulated power converter in parallel with an output of one or more additional pulse-width modulated power converters.
     
    12. The method of claim 1, the method further comprising a step of connecting only the input of the pulse-width modulated power converter in parallel with an input of one or more additional pulse-width modulated power converters.
     
    13. A system for measuring an integrity of a pulse-width modulated power converter (14), the system comprising:

    the pulse-width modulated power converter (14);

    a first current sensor (42) which senses current on a power converter input that includes an alternating component and a ripple component;

    a second current sensor (46) which senses current on a power converter output that includes an alternating component and a ripple component; and

    a programmable controller having loaded therein an integrity measuring algorithm, wherein the integrity measuring algorithm analyzes data from the first and second current sensors, in a way that either:

    a) a ripple component of the input current (Iin), a ripple component of the output current (Iout) and an alternating component of the output current (Iout); or

    b) a ripple component of the input current (Iin), a ripple component of the output current (Iout), an alternating component of the output voltage (Vout) and the power factor average;

    are processed to calculate an equivalent series resistance of the power converter, and as a result of the analysis, indicates an impending failure of the power converter.


     


    Ansprüche

    1. Verfahren zum Messen der Integrität eines pulsbreitenmodulierten Leistungswandlers (14), wobei das Verfahren die folgenden Schritte umfasst:

    Überwachen von Eingangsdaten, die mit einem Eingang zum pulsbreitenmodulierten Leistungswandler (14) assoziiert sind;

    Überwachen von Ausgangsdaten, die mit einem Ausgang vom pulsbreitenmodulierten Leistungswandler (14) assoziiert sind; und

    Korrelieren der Eingangsdaten und der Ausgangsdaten, um eine Verschlechterung der Leistung des Leistungswandlers zu erkennen, so dass entweder:

    a) eine Welligkeitskomponente des Eingangsstroms (Iin), eine Welligkeitskomponente des Ausgangsstroms (Iout) und eine alternierende Komponente des Ausgangsstroms (Iout); oder

    b) eine Welligkeitskomponente des Eingangsstroms (Iin), eine Welligkeitskomponente des Ausgangsstroms (Iout), eine alternierende Komponente der Ausgangsspannung (Vout) und der Leistungsfaktordurchschnitt;
    verarbeitet werden, um einen äquivalenten Reihenwiderstand des Leistungswandlers zu berechnen.


     
    2. Verfahren nach Anspruch 1, wobei der Korrelationsschritt ein Bestimmen der Verschlechterung der Leistung der korrelierten Eingangs- und Ausgangsdaten einschließt, wodurch ein bevorstehender Ausfall des Leistungswandlers (14) angezeigt wird.
     
    3. Verfahren nach Anspruch 2, ferner umfassend einen Schritt einer Initiierung einer automatischen Wartungsaktivität infolge der Anzeige des bevorstehenden Ausfalls, wodurch ein Ausfall des Leistungswandlers abgewendet wird, während der Leistungswandler (14) Leistung an das Stromversorgungsnetz (30) ausgibt.
     
    4. Verfahren nach Anspruch 2, wobei eine Anomalie eine wesentliche Änderung eines äquivalenten Reihenwiderstands des Leistungswandlers ist, wodurch der bevorstehende Ausfall einer Komponente des Leistungswandlers angezeigt wird.
     
    5. Verfahren nach Anspruch 1, wobei der Korrelationsschritt durchgeführt wird, während der Leistungswandler (14) betriebsfähig ist, und wobei der Korrelationsschritt ein Anzeigen eines bevorstehenden Ausfalls des Leistungswandlers (14) und ein Berechnen eines äquivalenten Reihenwiderstands für den Leistungswandler einschließt.
     
    6. Verfahren nach Anspruch 5, wobei der Korrelationsschritt ein Aufzeichnen des äquivalenten Reihenwiderstands als eine Funktion der Zeit einschließt.
     
    7. Verfahren nach Anspruch 5, wobei das Berechnen des äquivalenten Reihenwiderstands ein Kompensieren von wenigstens einem von Temperaturänderungen, Leistungsfaktoränderungen und Ausgangsspannungsänderungen einschließt.
     
    8. Verfahren nach Anspruch 5, wobei das Berechnen des äquivalenten Reihenwiderstands eine Mittelwertbildung eines Probe-für-Probe-Produkts von, einer Welligkeit eines Eingangsstroms, einer ersten Ableitung einer Welligkeit eines Ausgangsstroms und einer alternierenden Komponente von einem von dem Ausgangsstrom und einer Ausgangsspannung einschließt.
     
    9. Verfahren nach Anspruch 1, wobei die Überwachungsschritte ein Überwachen eines Eingangsstroms und eines Ausgangsstroms einschließen, und wobei die Korrelationsschritte ein Berechnen eines äquivalenten Reihenwiderstands aus den Werten der Eingangs- und Ausgangsströme einschließen.
     
    10. Verfahren nach Anspruch 1, wobei die Überwachungsschritte ein Überwachen eines Eingangsstroms, eines Ausgangsstroms und einer Ausgangsspannung einschließen, und wobei der Korrelationsschritt ein Berechnen eines äquivalenten Reihenwiderstands aus den Werten des Eingangsstroms, des Ausgangsstroms und der Ausgangsspannung einschließt.
     
    11. Verfahren nach Anspruch 1, das Verfahren ferner umfassend einen Schritt des Verbindens von nur dem Ausgang des pulsbreitenmodulierten Leistungswandlers parallel zu einem Ausgang eines oder mehrerer zusätzlicher pulsbreitenmodulierter Leistungswandler.
     
    12. Verfahren nach Anspruch 1, das Verfahren ferner umfassend einen Schritt des Verbindens von nur dem Eingang des pulsbreitenmodulierten Leistungswandlers parallel zu einem Eingang eines oder mehrerer zusätzlicher pulsbreitenmodulierter Leistungswandler.
     
    13. System zum Messen einer Integrität eines pulsbreitenmodulierten Leistungswandlers (14), das System umfassend:

    den pulsbreitenmodulierten Leistungswandler (14);

    einen ersten Stromsensor (42), der Strom am Leistungswandlereingang erfasst, der eine alternierende Komponente und eine Welligkeitskomponente einschließt;

    einen zweiten Stromsensor (46), der Strom am Leistungswandlerausgang erfasst, der eine alternierende Komponente und eine Welligkeitskomponente einschließt; und

    eine programmierbare Steuerung, auf der ein Integritätsmessalgorithmus geladen ist, wobei der Integritätsmessalgorithmus Daten vom ersten und zweiten Stromsensor analysiert, so dass entweder:

    a) eine Welligkeitskomponente des Eingangsstroms (Iin), eine Welligkeitskomponente des Ausgangsstroms (Iout) und eine alternierende Komponente des Ausgangsstroms (Iout); oder

    b) eine Welligkeitskomponente des Eingangsstroms (Iin), eine Welligkeitskomponente des Ausgangsstroms (Iout), eine alternierende Komponente der Ausgangsspannung (Vout) und der Leistungsfaktordurchschnitt;
    verarbeitet werden, um den äquivalenten Reihenwiderstand des Leistungswandlers zu berechnen, und infolge der Analyse ein bevorstehender Ausfall des Leistungswandlers angezeigt wird.


     


    Revendications

    1. Méthode de mesure de l'intégrité d'un convertisseur de puissance à modulation de largeur d'impulsion (14), comprenant les étapes suivantes :

    contrôle des données d'entrée associées à une entrée du convertisseur de puissance à modulation de largeur d'impulsion (14) ;

    contrôle des données de sortie associées à une sortie du convertisseur de puissance à modulation de largeur d'impulsion (14) ; et

    corrélation entre les données d'entrée et les données de sortie pour identifier une dégradation de la performance du convertisseur de puissance de telle façon que :

    a) soit un composant ondulatoire du courant d'entrée (lin), un composant ondulatoire du courant de sortie (Iout) et un composant en alternance du courant de sortie (Iout) ;

    b) soit un composant ondulatoire du courant d'entrée (lin), un composant ondulatoire du courant de sortie (Iout), un composant en alternance de la tension de sortie (Iout) et la moyenne du facteur de puissance ;

    soient traités pour permettre de calculer une résistance en série équivalente du convertisseur de puissance.


     
    2. La méthode de la revendication 1, selon laquelle l'étape de corrélation inclut la détermination d'une dégradation de la performance dans les données d'entrée et de sortie mises en corrélation, indiquant ainsi une défaillance prochaine du convertisseur de puissance (14).
     
    3. La méthode de la revendication 2, comprenant également une étape de lancement d'une activité de maintenance automatisée en conséquence de l'indication de défaillance prochaine, évitant ainsi une défaillance du convertisseur de puissance pendant que le convertisseur de puissance (14) transmet de la puissance au réseau électrique (30).
     
    4. La méthode de la revendication 2, selon laquelle une anomalie est un changement substantiel dans une résistance en série équivalente du convertisseur de puissance, indiquant ainsi une défaillance prochaine d'un composant du convertisseur de puissance.
     
    5. La méthode de la revendication 1, selon laquelle l'étape de corrélation est exécutée pendant que le convertisseur de puissance (14) est opérationnel, et selon laquelle l'étape de corrélation inclut l'indication de défaillances prochaines du convertisseur de puissance (14) et le calcul d'une résistance en série équivalente pour le convertisseur de puissance (14).
     
    6. La méthode de la revendication 5, selon laquelle l'étape de corrélation inclut l'enregistrement de la résistance en série équivalente en fonction du temps.
     
    7. La méthode de la revendication 5, selon laquelle le calcul de la résistance en série équivalente inclut la compensation d'au moins un des changements de température, un des changements du facteur de puissance et un des changements de la tension de sortie.
     
    8. La méthode de la revendication 5, selon laquelle le calcul de la résistance en série équivalente inclut la détermination de la moyenne d'un produit ondulatoire échantillon par échantillon du courant d'entrée, d'un premier dérivé ondulatoire du courant de sortie et d'un composant en alternance d'un courant de sortie et d'une tension de sortie.
     
    9. La méthode de la revendication 1, selon laquelle les étapes de contrôle incluent le contrôle d'un courant d'entrée et d'un courant de sortie, et selon laquelle l'étape de corrélation inclut le calcul d'une résistance en série équivalente à partir des valeurs des courants d'entrée et de sortie.
     
    10. La méthode de la revendication 1, selon laquelle les étapes de contrôle incluent le contrôle d'un courant d'entrée, d'un courant de sortie et d'une tension de sortie, et selon laquelle l'étape de corrélation inclut le calcul d'une résistance en série équivalente à partir des valeurs du courant d'entrée, du courant de sortie et de la tension de sortie.
     
    11. La méthode de la revendication 1, cette méthode comprenant également une étape consistant à connecter seulement la sortie du convertisseur de puissance à modulation de largeur d'impulsion en parallèle avec une sortie d'un ou de plusieurs convertisseurs de puissance à modulation de largeur d'impulsion additionnels.
     
    12. La méthode de la revendication 1, cette méthode comprenant également une étape consistant à connecter seulement l'entrée du convertisseur de puissance à modulation de largeur d'impulsion en parallèle avec une entrée d'un ou de plusieurs convertisseurs de puissance à modulation de largeur d'impulsion additionnels.
     
    13. Système pour mesurer une intégrité d'un convertisseur de puissance à modulation de largeur d'impulsion, le système comprenant :

    le convertisseur de puissance à modulation de largeur d'impulsion (14) ;

    un premier détecteur de courant (42) qui détecte du courant sur une entrée du convertisseur de puissance incluant un composant en alternance et un composant ondulatoire ;

    un deuxième détecteur de courant (46) qui détecte du courant sur une sortie du convertisseur de puissance incluant un composant en alternance et un composant ondulatoire ; et

    un contrôleur programmable y ayant chargé un algorithme de mesure de l'intégrité, dans lequel l'algorithme de mesure de l'intégrité analyse les données des premier et deuxième détecteurs de courant, de telle façon que :

    a) soit un composant ondulatoire du courant d'entrée (lin), un composant ondulatoire du courant de sortie (Iout) et un composant en alternance du courant de sortie (Iout) ;

    b) soit un composant ondulatoire du courant d'entrée (lin), un composant ondulatoire du courant de sortie (Iout), un composant en alternance de la tension de sortie (Iout) et la moyenne du facteur de puissance ;

    soient traités pour permettre de calculer une résistance en série équivalente du convertisseur de puissance, et, en conséquence de l'analyse, indiquent une défaillance prochaine du convertisseur de puissance.


     




    Drawing





























    Cited references

    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