(19)
(11)EP 3 579 007 A1

(12)EUROPEAN PATENT APPLICATION

(43)Date of publication:
11.12.2019 Bulletin 2019/50

(21)Application number: 18176500.9

(22)Date of filing:  07.06.2018
(51)International Patent Classification (IPC): 
G01R 31/36(2019.01)
H01M 10/48(2006.01)
(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
Designated Extension States:
BA ME
Designated Validation States:
KH MA MD TN

(71)Applicant: Siemens Aktiengesellschaft
80333 München (DE)

(72)Inventors:
  • DEBRECENI, Tibor
    1125 Budapest 6 (HU)
  • BALAZS, Gergely György
    1034 Budapest (HU)
  • SZABO, Peter
    1104 Budapest (HU)

  


(54)METHOD AND APPARATUS FOR ESTIMATING A STATE OF CHARGE OF A BATTERY


(57) According to the proposed embodiments a state of charge of a battery is estimated by a battery model specific to the battery, the battery model providing a section-wise defined correlation of terminal voltage values depending on state of charge values wherein each of said sections of the battery model delimits a monotonic dependence of the correlation from others of said sections. By segmenting the correlation of terminal voltage values depending on a state of charge value into sections, each segment within such the section-wise defined correlation of terminal voltage values depending on state of charge values is mathematically spoken a bi-unique function suitable for transformation into an inverse function defined within the section.
According to a further embodiment, the estimated state of charge is continuously refined by an iterative feedback loop including a coulomb counting for estimating a battery charge value, wherein refined estimated battery charge values state of charge values at a previous cycle are projected forward to the current cycle.




Description

TECHNICAL FIELD



[0001] The disclosed embodiments generally relate a method and apparatus for estimating a state of charge of a rechargeable battery.

BACKGROUND



[0002] Rechargeable batteries have become increasingly important mainly to supply energy storages in stationary or e-mobility applications, i.e. electric vehicles such as cars and aircrafts, or electronic equipment such as mobile devices.

[0003] A state of charge value or SoC value indicates a usable capacity of the battery. This value is usually provided in percentage of a full capacity of the battery for maintenance personnel or an end user such as driver or a pilot. The state of charge cannot be measured directly as a quantity, it is rather estimated instead. As part of an energy storage system, a Battery Management System or BMS is usually responsible for the estimation of the state of charge, which is crucial to be sufficient and reliable in safety- and life-critical applications.

[0004] Continuous monitoring of the state of charge of batteries by means which are as accurate as possible is necessary for users in many applications to avoid a stoppage or an interruption of the operation of the equipment served by such batteries. Existing battery cell capacities can only be fully utilized by state-of-charge monitoring.

[0005] Devices for estimating the state of charge of batteries are known in the art. The state of charge estimation may be provided by the Battery Management System using data from voltage measurements of each battery cell potential and a load or charge current measurements imposed on the battery or a battery pack.

[0006] According to a method heretofore known, a terminal voltage value - i.e. a voltage of a battery while loads connected to the battery - is measured, an upper voltage limit defining the end of the battery charging cycle and a lower voltage limit defining the end of the battery discharging cycle. Between these ends, a non-linear behavior of the battery is considered by an estimation algorithm applying a non-linear battery model as a correlation - or mathematically: a function - of a voltage value over the state of charge. Such battery model used for estimating the state of charge of a battery in operation may be provided beforehand using for example, empirical methods of experimentally obtaining a function of an open circuit voltage value - i.e. the voltage of a battery without load - of a reference battery depending on its state of charge. The state of charge of a battery to be estimated may then be inferred by measuring the open circuit voltage and determining a respective state of charge according to the battery model. Instead of determining the open circuit voltage of the battery under regular conditions, a normal discharge process may be done, thereby adding a correction to the measured terminal voltage in order to determine open circuit voltage characteristics.

[0007] A major drawback of this method heretofore known is caused by an indeterminacy in inferring the state of charge derived from the non-linear function by a currently measured voltage value, as the function of the voltage value over the state of charge provided by the battery model may not always return a single state of charge value for a given voltage value. In other words, the function of the voltage value over the state of charge may return more than one state of charge value for a given voltage value.

[0008] Accordingly there is a need in the art for uniquely determining a state of charge estimation of a battery using a battery model for a given voltage value.

SUMMARY



[0009] Embodiments for estimating a state of charge as described herein generally involve applying a battery model for estimating a state of charge of the battery by measuring at least a terminal voltage of the battery.

[0010] In one embodiment a method for estimating a state of charge of a battery is disclosed, including the steps of:
  1. a) providing a battery model specific to the battery, the battery model providing a section-wise defined correlation of terminal voltage values depending on state of charge values wherein each of said sections of the battery model delimits a monotonic dependence of the correlation from others of said sections;
  2. b) measuring a terminal voltage value of the battery;
  3. c) identifying one of said sections of the battery model based on at least one operational condition of the battery;
  4. d) retrieving, within the identified section of the battery model, a state of charge value correlating to the measured terminal voltage, and;
  5. e) returning the retrieved state of charge value as an estimated state of charge of the battery.


[0011] The proposed method uses a correlation - or mathematically: a function - included in a battery model wherein, according to the correlation, each terminal voltage value is dependent on a state of charge value.

[0012] As the terminal voltage is known by a measurement in step b), the inverse correlation - or mathematically: inverse function - has to be determined, i.e. the state of charge value depending on the measured terminal voltage value. However, the inverse correlation of the voltage value over the state of charge value is not biunique as it may return more than one state of charge value for a given voltage value.

[0013] Mathematically expressed, a transformation into an inverse function is not possible for a function in which the curve progression of the dependent value - the voltage value - is not strictly monotonically increasing or decreasing. In fact a typical curve progression of the terminal voltage value as a function of the state of charge value is not monotonically decreasing, but in some parts forms a constant progression or even an increasing curve progression as illustrated in Fig. 2 of the drawing. This leads to a situation where the inverse function produces more than one state of charge value for a given voltage value, inhibiting a unique determination of the state of charge value for a given voltage value.

[0014] According to an embodiment of the invention this problem is resolved by an approach of segmenting the correlation of terminal voltage values depending on a state of charge value into sections, wherein the correlation exhibits a monotonic dependence. A monotonic dependence is either a monotonic increasing or a monotonic decreasing of the curve progression.

[0015] Each function segment within such a section-wise defined function of terminal voltage values depending on state of charge values is mathematically spoken a biunique function suitable for transformation into an inverse function defined within the section.

[0016] This means that the section-wise defined inverse correlation of the state of charge values depending on the known terminal voltage values produces exactly one state of charge value for one given terminal voltage value.

[0017] Applying said section-wise defined correlation within the battery model the proposed method provides a step of measuring the terminal voltage value of the battery, followed by a step of identifying the suitable section of the battery model, wherein the identification of the section is based on at least one operational condition of the battery. Subsequently, a state of charge value correlating to the measured terminal voltage is determined within the identified section of the battery model. Finally, the determined state of charge value is returned as an estimated state of charge of the battery.

[0018] According to an embodiment, the estimated state of charge is refined by a correction factor imposed to the determined state of charge. This correction factor is determined by coulometry principle wherein the quantity of energy delivered to or by the battery are determined and wherein a value of the remaining battery charge available for delivery is obtained from the difference between a nominal battery capacity and the energy delivered by the battery. The remaining battery charge so obtained is used for correcting the estimated state of charge.

[0019] According to a preferred embodiment, an iterative feedback loop of estimating the state of charge is applied, wherein the battery capacity and the estimated state of charge is refined at each time interval or cycle by operating as an iterative feedback loop projecting forward the refined battery capacity at the previous cycle to the current cycle and using the refined battery capacity at the previous cycle as one of the operational conditions for identifying one of said sections of the battery model.

[0020] According to an embodiment, the terminal voltage values of the battery model are expressed by equivalent open circuit voltage values, wherein the equivalent open circuit voltage values are compensated by a voltage drop in an internal impedance of the battery.

[0021] According to an embodiment, the operational conditions of the battery include one or more of a current load on the battery, a deemed open circuit voltage of the battery, an internal resistance or impedance of the battery and/or a temperature of the battery.

[0022] In another embodiment a battery management system for estimating a state of charge of a battery is disclosed, the battery management system including a battery model, a measuring module, an indexing module, a correlation module and an output module. Said modules are carrying out the method for estimating the state of charge according to the embodiments as described above.

DESCRIPTION OF THE DRAWING



[0023] The objects as well as further advantages of the present embodiments will become more apparent and readily appreciated from the following description of the preferred embodiments, taken in conjunction with the accompanying drawing in which:
FIG. 1
shows a section-wise defined correlation of a voltage of a battery as a function of the state of charge of the battery as used in a battery model according to an embodiment;
FIG. 2
shows a graph of a voltage of a battery as a function of the state of charge of the battery according to the state of the art; and;
FIG. 3
shows a diagram illustrating an operational flowchart according to an embodiment.

DETAILED DESCRIPTION



[0024] Fig. 2 shows a graph of an open circuit voltage of a battery as a correlation or function of the state of charge of the battery whereby parameters of this correlation may be stored in a battery model and used by a battery management system according to the state of the art.

[0025] The curve depicted in FIG. 2 shows an exemplary correlation of a battery voltage in dependency of its state of charge for a given battery operating at a pre-determined temperature. This correlation has been pre-emptively determined for a particular battery type and the parameter values of the correlation have been stored in a parameter value resource - optionally a lookup table - of a battery model modelling the behavior of the battery according to the curve depicted in FIG. 2.

[0026] The pre-emptive determination of the parameter values may have been carried out either experimentally, derived empirically or semi-empirically based on tests of standard cells, by a deductive approach considering chemical and physical properties of the battery, or by a combination of said determination approaches. According to the exemplary correlation, a voltage value V on the ordinate of the coordinate system shown in FIG. 2 is dependent on a state of charge value S on the abscissa.

[0027] This method - which is also referred to as Open Circuit Voltage-based (OCV) method - is based on the voltage measurement of a battery potential. Every battery technology has its own, distinct voltage characteristic, which has many dependencies. The most conspicuous dependency is the open circuit voltage, i.e. the voltage of a battery without load, depending on a state of charge of the battery. By measurements, this characteristic can be obtained. Measuring the voltage of the battery without load, the state of charge can be determined with good accuracy. A major drawback of this method is that it can be used only with batteries in an idle operation and not during discharging or charging the battery, since the impedance of the battery alters the voltage significantly with the load current. Another drawback in using this method is a property of the voltage characteristics which curve progression is not monotonically decreasing. This property will be addressed hereinafter further below.

[0028] After the pre-emptive determination of the correlation of a battery voltage in dependency of its state of charge for a given battery has been stored has been finished, the parameter values of the correlation are stored in a parameter value resource of a battery model. The battery model now serves for an estimation of the state of charge.

[0029] As the open circuit voltage value V may be measured, the inverse correlation - or mathematically: inverse function - has to be determined, i.e. the state of charge value S depending on the measured terminal voltage value V. However, the inverse correlation of the voltage value over the state of charge value is not biunique as it returns more than one state of charge value S1, S2 for a at least one voltage value V1 as depicted in the drawing. The reason for that is that the curve progression of the terminal voltage value V as a function of the state of charge value S is not monotonically decreasing, but in some parts forms an increasing curve progression. This leads to a situation where the inverse function produces more than one state of charge value S1,S2 for a given voltage value V1, inhibiting a unique determination of the state of charge value for this voltage value V1.

[0030] Methods for a state of charge estimation using a battery model encounter a number of limitations which arise from the necessity of determining the inverse correlation:
  • Certain types of batteries have a behavior - not shown in FIG. 2 - in that there are more state of charge values belonging to the same voltage value in some circumstances. In the case of Lithium-ion or Li-Ion batteries, for example, when operated at low - particularly temperatures below 0° C - temperatures, a unique determination of the state of charge value for a measured voltage value is not possible. In other words: the inverse correlation of the voltage value over the state of charge value is not biunique for Li-Ion batteries operated at temperatures below 0° C. However, Li-Ion batteries operated at room temperature allow for a unique determination of the state of charge value for a measured voltage value. In other words: the inverse correlation of the voltage value over the state of charge value is biunique for Li-Ion batteries operated at room temperature. Other examples of batteries which allow for a unique determination of the state of charge value for a measured voltage value only in certain temperature ranges include LiFePo, LiPo, NiMH, NiZn, Lead-acid and NiCd batteries. In the latter case of NiCd batteries a unique determination is dependent of an absence of overcharging effects, which means that a unique determination is not possible when overcharging effects occur or have occurred.
  • Other types of batteries, e.g. lithium-sulfur or Li-S batteries, have a behavior - shown in FIG. 2 - where there are more state of charge values belonging to the same voltage value in all circumstances. In other words: the inverse correlation of the voltage value over the state of charge value is not biunique for Li-S batteries for all conditions.


[0031] Known methods of estimating the state of charge are partially applicable for widely used lithium-based batteries, for example Li-Ion batteries, lithium-polymer (Li-Po) batteries, or lithium-iron-phosphate (Li-Fe-Po) batteries, because their behavior is such that there is one definite state of charge value belonging to one voltage level in some circumstances, meaning the inverse correlation - or mathematically: inverse function - of the discharge voltage characteristic of the cell is a function of state of charge or, alternatively, the Depth of Discharge (DoD).

[0032] Apart from the model-based method shown above, further methods of estimating the state of charge of a battery are known in the art, including a Coulomb-Counting (CC) method. The Coulomb-Counting is based on continuous measurements of discharge and/or charge currents on the battery. The current values are integrated over time in order to determine the already used capacity of the battery. Comparing this value to the nominal capacity, the state of charge can be estimated. A major drawback of this method is a drift error in the estimated capacity which is caused by a constantly accumulated deviation due to the integration.

[0033] FIG. 1 shows a section-wise defined correlation of a voltage of a battery as a function of the state of charge of the battery as used in a battery model according to an embodiment.

[0034] This embodiment is essentially based on the idea of splitting the curve of the voltage characteristics of the battery known from FIG. 2 into a multiplicity of sections 1,2,3,...M-1, M, as a function of a state of charge. The sections 1,2,3,...M-1, M are symbolized by an encircled ordinal number in the drawing.

[0035] Unlike the example shown in conjunction with FIG. 2, the embodiment according to FIG. 1 uses an equivalent open circuit voltage EOV as a unit on the ordinate of the depicted coordinate system. The resulting characteristic is hereinafter called »Equivalent Open Circuit Voltage Characteristic« or EOCVC.

[0036] Equivalent open circuit means, that the voltage characteristic of the battery is compensated in order to be independent from any currents, which may be due to impedance. The terminal voltage values of the battery model, in other words are expressed by equivalent open circuit voltage values EOV, wherein the equivalent open circuit voltage values EOV are compensated by a voltage drop in an internal impedance of the battery caused by a current drained off the battery.

[0037] Splitting the Equivalent Open Circuit Voltage Characteristic results in a number M of sections 1,2,3,...M-1, M wherein in each section the correlation exhibits a monotonic dependence. According to the exemplary curve shown in FIG. 1, the curve is strictly monotonic decreasing in sections 1,2,4, M-1 and M and strictly monotonic increasing in section 5. By applying mathematical derivation methods, such as dEOV/dS or derivations of a higher order a sufficient number of sections can be determined. It can be stated, that the more sections can be distinguished, the more accurate the estimation of the state of charge will be.

[0038] Each function segment within such a section-wise defined function of equivalent open circuit voltage values EOV depending on state of charge values S is mathematically spoken a biunique function suitable for transformation into an inverse function defined within the section 1,2,3,...M-1,M.

[0039] This means that the section-wise defined inverse correlation of the state of charge values S depending on the known open circuit voltage values EOV produces exactly one state of charge value for one given terminal voltage value within the actually selected section 1,2,3,...M-1,M.

[0040] Applying said section-wise defined correlation within the battery model the embodiment provides a step of measuring the terminal voltage value or the open circuit voltage values EOV of the battery, followed by a step of identifying the suitable section 1,2,3,...M-1, M of the battery model, wherein the identification of the section 1,2,3,...M-1,M is based on at least one operational condition of the battery. Subsequently, a state of charge value S correlating to the measured terminal voltage is determined within the identified section 1,2,3,...M-1, M of the battery model.

[0041] Of course, the section-wise defined correlation within the battery model is beneficial for both types of batteries as stated above, including both:
  • types of batteries which have a behavior in that there are more state of charge values belonging to the same voltage value in some circumstances, e.g. Li-Ion batteries; and;
  • types of batteries which have a behavior in that there are more state of charge values belonging to the same voltage value in all circumstances, e.g. Li-S batteries.


[0042] The benefits of the section-wise defined correlation are due to the fact that operational circumstances are not to be regarded for applying the inverse correlation. These circumstances are advantageously considered within the battery model.

[0043] FIG. 3 shows a diagram illustrating an operational flowchart according to a further embodiment. According to this embodiment, the estimated state of charge is continuously refined by an iterative feedback loop CYC including a measurement of parameters and retrieving of parameters within the battery model, wherein refined estimated state of charge values at a previous cycle are projected forward to the current cycle.

[0044] Within a first section S1 of the operational flowchart, a step 101 of measuring a terminal voltage value of the battery, a step 102 of measuring a current drained of the battery and a step 103 of determining a battery capacity is provided. The battery capacity - which usually expresses a difference between a nominal capacity of the battery and the hitherto used capacity - used in the current cycle of the iterative feedback loop CYC was determined in a previous cycle.

[0045] Within a second section S2 of the operational flowchart, a step 201 of inputting the terminal voltage value and the current value into a battery model is provided. An equivalent open circuit voltage value 402 is determined by a difference of the terminal voltage value of the battery measured in step 101 and the voltage drop in a known internal impedance caused by the current value drained off the battery measured in step 102. According to an embodiment, the known internal impedance and/or resistance is included in the battery model and retrieved therefrom.

[0046] Within a third section S3 of the operational flowchart, a sequence of identifying a section within the battery model based on at least one operational condition of the battery is provided. The operational condition may include a battery capacity of a previous cycle in order to coarsely identify the location on an abscissa of the correlation represented within the battery model, thereby uniquely identifying the section on the state of charge portion of the abscissa.

[0047] In a first decision step 301, a first decision is made of whether the first section within the battery model is applicable. If the first section is applicable, which is represented by a branch Y (»Yes«) pointing vertically downward from decision step 301, a subsequent step 401 is carried out. If the first section is not applicable, represented by a branch N (»No«) pointing horizontally to the right from decision step 301, a subsequent decision step 302 is carried out. In the subsequent second decision step 302, a second decision is made of whether the second section within the battery model is applicable. If the second section is applicable, which is represented by a branch Y (»Yes«) pointing vertically downward from the second decision step 302, the subsequent step 401 is carried out. If the second section is not applicable, represented by a branch N (»No«) pointing horizontally to the right from the second decision step 302, a subsequent - not shown - decision step between decision step 302 and decision step 303 is carried out. The sequence of decision steps continues - symbolized by three dots in the drawing - until an ordinal number equal to the number M of sections of the battery model has been reached. In the final (M-1)th decision step 303, a decision is made of whether the (M-1)th section within the battery model is applicable. Regardless of whether the (M-1)th section is applicable or not - which is represented by a branch Y (»Yes«) and a branch N (»No«), both pointing vertically downward from the final decision step 303 - the subsequent step 401 is carried out.

[0048] Within a fourth section S4 of the operational flowchart, the step of indexing 401 by an indexing module returns the identified section number within the battery model. In step 403, the state of charge value 404 correlating to the measured terminal voltage is retrieved within the identified section of the battery model, thereby using the equivalent open circuit voltage value 402. This state of charge value retrieved within the identified section of the battery model is returned as estimated state of charge 404 of the battery.

[0049] In a coulomb counting step 406 a battery capacity 405 determined in the current cycle of the iterative feedback loop is calculated using a differentially used capacity - i.e. a capacity used in the currently operated cycle - by integrating over time the current value measured by step 102 and the battery capacity 103 which was determined in the previous cycle and delivered by step 103. The calculated battery capacity 405 is the battery capacity 405 in the current cycle. A refined state of charge, derived as a quotient of the calculated battery capacity 405 and the nominal battery capacity, is returned by step 605.

[0050] By a calculation step 407 - symbolized by a crossed circle in the drawing - the determined state of charge derived by the battery charge 405 in the current cycle and the estimated state of charge 404 are used to calculate a correction factor 501. According to an embodiment, the correction factor 501 is calculated as a difference between the estimated state of charge 404 and the determined state of charge derived by the battery charge 405 in the current cycle, as symbolized by the »+« and »-« symbols in the drawing.

[0051] Within a fifth section S5 of the operational flowchart, the correction factor 501 is fed to a refining step 502 for refining the battery capacity 503 delivered by step 103 from the previous cycle by refining the battery capacity 503 from the previous cycle with the correction factor 501. According to an embodiment, the refining step 502 includes the calculation of

wherein:
QB(N)
denotes the refined battery capacity determined within the current cycle N;
QB(N-1)
denotes the battery capacity 503 delivered by step 103 from the previous cycle N-1;
CF
denotes the correction factor 501;


[0052] The refined battery capacity determined within the current cycle is returned by step 604. This refined battery capacity determined within the current cycle is returned by step 604 to the subsequent cycle CYC wherein it will be used
  • equivalently to the former step 103 - as a new value for the battery capacity in the subsequent circle. In the index counting used in the formula above, this subsequent cycle would have an index of N+1. In other words, the step of refinement is updating the value of the nominal capacity, which and that is looped back to the Coulomb Counter 406 in the next step.


[0053] Within a sixth section S6 of the operational flowchart, a step 601 of triggering new measurements steps 602,603 and repeating the cycle by a feedback loop CYC is carried out. These new measurement steps include a step 602 of measuring the terminal voltage value of the battery and the step 603 of measuring the actual current drained of the battery. The steps 602,603 are equivalently used as steps 101,102 in the preceding cycle.

[0054] The steps described above are preferably executed by a battery management system interfacing a digital and/or analog data representation of a battery model 201, a - not shown - measuring module configured to carry out the steps 101,102,602,603, a - not shown - indexing module operable to carry out the steps 301,302,303,401, a - not shown - correlation module operable to carry out the step 403 and an output module to carry out the step 604.

[0055] According to the proposed embodiments a state of charge of a battery is estimated by a battery model specific to the battery, the battery model providing a section-wise defined correlation of terminal voltage values depending on state of charge values wherein each of said sections of the battery model delimits a monotonic dependence of the correlation from others of said sections. By segmenting the correlation of terminal voltage values depending on a state of charge value into sections, each segment within such the section-wise defined correlation of terminal voltage values depending on state of charge values is mathematically spoken a bi-unique function suitable for transformation into an inverse function defined within the section.

[0056] According to an embodiment, the estimated state of charge is continuously refined by an iterative feedback loop including a coulomb counting for estimating a battery charge value, wherein refined estimated battery charge values state of charge values at a previous cycle are projected forward to the current cycle.

[0057] According to an embodiment, a - not shown - iterative feedback loop execution unit is executing an iterative feedback loop CYC projecting forward the battery charge at a previous cycle to the current cycle and using a correction factor in order to return a refined estimated state of charge of the battery.

[0058] According to an embodiment, the steps 407, 502 of calculating the correction factor and refining the estimated state of charge by the correction factor are executed by a - not shown - correction module.

[0059] According to an embodiment, the coulomb counting step 406 is executed by a - not shown - coulomb counting module.

[0060] The proposed embodiments allow for an easy implementation in a battery management system due to the fact that complex computation time and power demanding operations are not necessary. The embodiments provide reliable state of charge values even in implementations where a signal-to-noise-ratio (SNR) of the voltage and current measurements is of a rather low quality.

[0061] As the embodiments solely use simple mathematical operations along with scalar values - instead of using matrix operations like in Kalman filters - an implementation in real-time systems using standard microcontroller-based battery management systems is suitable and advantageously results in low costs for development.

[0062] The embodiments are of an adaptable and scalable nature. Merely by changing certain parts modules - particularly the battery model - the modules may be re-used for all kinds of battery technologies. Even the implemented algorithm is independent from voltage and current levels, so that it can re-used for battery cells, battery packs or modules as well.

[0063] In order to customize to a specific application with a specific battery cell, only a parameterization update is needed. Changing the battery technology only requires a parameter table update within the battery model.

[0064] It is to be understood that the elements and features recited in the appended claims may be combined in different ways to produce new claims that likewise fall within the scope of the present invention. Thus, whereas the dependent claims appended below depend from only a single independent or dependent claim, it is to be understood that these dependent claims can, alternatively, be made to depend in the alternative from any preceding or following claim, whether independent or dependent, and that such new combinations are to be understood as forming a part of the present specification.

[0065] While the present invention has been described above by reference to various embodiments, it should be understood that many changes and modifications can be made to the described embodiments. It is therefore intended that the foregoing description be regarded as illustrative rather than limiting, and that it be understood that all equivalents and/or combinations of embodiments are intended to be included in this description.


Claims

1. A method for estimating a state of charge of a battery, including the steps of:

a) providing a battery model specific to the battery, the battery model providing a section-wise defined correlation of terminal voltage values depending on state of charge values wherein each of said sections of the battery model delimits a monotonic dependence of the correlation from others of said sections;

b) measuring a terminal voltage value of the battery;

c) identifying one of said sections of the battery model based on at least one operational condition of the battery;

d) retrieving, within the identified section of the battery model, a state of charge value correlating to the measured terminal voltage, and;

e) returning the retrieved state of charge value as an estimated state of charge of the battery.


 
2. The method according to claim 1, including the steps of:

- determining a used capacity within a current cycle by integrating a measured current drained of the battery over time within a current cycle;

- determining a battery capacity in the current cycle by using a battery capacity of a preceding cycle and the used capacity within the current cycle;

- calculating a correction factor using the battery capacity in the actual cycle and the state of charge value determined by step d); and;

- refining the battery capacity determined within the current cycle by multiplying the determined battery capacity with the correction factor.


 
3. The method according to claim 2, refining the battery capacity at each cycle by operating as an iterative feedback loop projecting forward the battery capacity at a previous cycle to the current cycle and using a correction factor in order to return a refined battery capacity.
 
4. The method according to one of the aforementioned claims, wherein the terminal voltage values of the battery model are expressed by equivalent open circuit voltage values, the equivalent open circuit voltage values being compensated by a voltage drop in an internal impedance of the battery.
 
5. The method according to one of the aforementioned claims, wherein the operational condition of the battery includes one or more of:

- a current load on the battery;

- a deemed open circuit voltage of the battery;

- an internal impedance and/or resistance of the battery;

- a temperature of the battery.


 
6. A battery management system for estimating a state of charge of a battery, including:

- a battery model specific to the battery, the battery model providing a section-wise defined correlation of terminal voltage values depending on state of charge values wherein each of said sections of the battery model delimits a monotonic dependence of the correlation from others of said sections;

- a measuring module configured to measure a terminal voltage value of the battery;

- an indexing module operable to identify one of said sections of the battery model based on an operational condition of the battery; and;

- a correlation module operable to retrieve, within the identified section of the battery model, a state of charge value correlating to the terminal voltage; and;

- an output module for returning the retrieved state of charge value as an estimated state of charge of the battery.


 
7. The battery management system according to claim 6 including
an iterative feedback loop execution unit for executing an iterative feedback loop CYC projecting forward a battery capacity value at a previous cycle to a current cycle and using a correction factor in order to return a refined estimated state of charge of the battery.
 




Drawing










Search report









Search report