Method for automatically estimating battery state of charge

Abstract

The automatic process for estimating the state of charge of a battery cell comprises a phase (116) of estimating the state of charge of the cell at an instant k based on an estimated change in entropy ΔS, said phase (116) comprising the following steps for a physical quantity selected from the group consisting of the internal temperature of the cell and the voltage across the terminals of the cell: calculating (118) an estimate of the physical parameter using an electrical model if the physical parameter is the voltage across the cell terminals, or using a thermal model if the physical parameter is the internal temperature, and calculating (122) the deviation of the estimate of the physical parameter from the measured value of the physical parameter; - The calculated deviation is used to construct (122) an estimate of the state of charge at the instant k.

Description

[Technical Field]

[0001] The present invention relates to a method for automatically estimating the state of charge of a battery, a recording medium for implementing this method, and an electronic battery management system. The present invention also relates to a vehicle incorporating this battery management system. [Background technology]

[0002] WO2020064959 describes a method for estimating the state of charge (SOC) of a battery from estimated values ​​of the entropy change (ΔS) and enthalpy change (ΔH) of the battery. More precisely, the state of charge (SOC) is estimated using the following relationship: SOC = α.ΔS + β.ΔH + γ, where α, β, and γ are parameters predetermined in a calibration phase. The enthalpy change (ΔH) of the battery is estimated using the following relationship: ΔH = -F.OCV - Ti.ΔS, where -OCV is the open circuit voltage of the battery cell. -Ti is the measured internal temperature of the battery cell. -F is the Faraday constant.

[0003] To estimate the open circuit voltage (OCV), WO2020064959 uses an electrical model of the battery. To estimate the entropy variation (ΔS), WO2020064959 uses a thermal model of the battery. The process described in WO2020064959 is advantageous in that it can be implemented during normal cell use and therefore in an electronic battery management system. Such battery management systems are more commonly known by the abbreviation BMS ("battery management system").

[0004] The inventors have realized that using the entropy variation ΔS to estimate the state of charge of a battery cell should make it possible to obtain a more accurate estimate of this state of charge, especially for battery cells whose open-circuit voltage changes little as a function of their state of charge. Such cells are used in particular in LFP ("lithium iron phosphate" or "lithium iron phosphate") batteries, since their open-circuit voltage changes little with their state of charge. Indeed, in LFP battery cells, the entropy variation ΔS of the cell varies significantly with its state of charge, making it possible to obtain greater accuracy in estimating its state of charge. However, in practice, this expected benefit has not been clearly achieved using the process described in WO2020064959.

[0005] The present invention aims to remedy this drawback by proposing an automatic process for estimating the state of charge of a battery cell that is more accurate than the process described in WO2020064959 while retaining its advantages.

[0006] The invention is set out in the accompanying claims. [Brief explanation of the drawings]

[0007] The invention will be better understood from reading the following description, given by way of non-limiting example only and made with reference to the drawings in which: [Figure 1] 1 is a partial schematic diagram of a vehicle equipped with an electric battery; [Figure 2] FIG. 2 is a schematic diagram of an electrical model of a battery cell in the vehicle of FIG. 1. [Figure 3] FIG. 2 is a schematic diagram of an estimator arrangement used to estimate the state of charge of the battery cells in the vehicle of FIG. 1. [Figure 4] 4 is a flowchart of a process for estimating the state of charge of a cell using the estimator shown in FIG. 3. [Figure 5]5 is a graph showing the state of charge estimation of a cell over time using the process shown in FIG. 4. DETAILED DESCRIPTION OF THE INVENTION

[0008] Chapter 1 of this document defines the terms and notation used in this document. Chapter 2 describes detailed examples of embodiments with reference to the drawings for the specific case where the cell for which the state of charge is to be estimated is a battery cell of an electric vehicle. Chapter 3 then introduces variations of these embodiments. Finally, Chapter 4 identifies the advantages of different embodiments.

[0009] Chapter 1: Terminology and Notation:

[0010] The same reference numerals are used in the figures to denote the same elements. In the remainder of this description, features and functions well known to those skilled in the art will not be described in detail.

[0011] In this description, "computing power" refers to the number of operations performed by an electronic computer. Reducing computing power therefore means reducing the number of operations required to achieve the same result or a result of the same nature.

[0012] The term "internal temperature" refers to the temperature inside the battery. If the temperature inside the battery is relatively uniform, the internal temperature is close to the temperature measurable at the outer surface of the battery cell. Therefore, the internal temperature also refers to the temperature of the outer surface of the battery.

[0013] In this description, the symbol " T " represents the mathematical transpose operation. The multiplication operation is represented by the operator "."

[0014] Chapter 2: Example of Implementation

[0015] FIG. 1 shows an electrically powered vehicle 2, better known as an "electric car." Electric vehicles are well known, and only the structural elements necessary to understand the remainder of this description are presented. Vehicle 2 is an electric motor 4 capable of rotating drive wheels 6 to drive the vehicle 2 on a road surface 8, and - The motor 4 is provided with a battery 10 for supplying electrical energy thereto.

[0016] The battery 10 has two electrical connection terminals 12, 14 and a number of electrical cells electrically connected between these terminals 12, 14. The terminals 12, 14 are connected to a supply electrical load, which in turn is connected to an electric motor 4.

[0017] For simplicity, only four electrical cells 18 to 21 are shown in Figure 1. Typically, these electrical cells are grouped into stages, which are connected in series between terminals 12, 14. Here, only two stages are shown. The first stage consists of cells 18 and 19, and the second stage consists of cells 20 and 21. Each stage has several branches connected in parallel. Each branch of a stage consists of one or more electrical cells connected in series. Here, the first stage consists of two branches, each consisting of a single electrical cell. The second stage is structurally identical to the first stage in the example shown in Figure 1.

[0018] It should be noted that all cells in battery 10 are structurally identical within manufacturing tolerances, and therefore only cell 18 will be described in detail herein.

[0019] Cell 18 has two electrical terminals 30, 32 that electrically connect to the other cells and to terminals 12 and 14 of battery 10. Cell 18 is also mechanically attached to the other cells of battery 10 without any degrees of freedom, forming what is often referred to as a cell "pack." Cell 18 can store electrical energy when not in use. This stored electrical energy is used to power motor 4, discharging cell 18. At other times, cell 18 can also receive and charge electrical energy.

[0020] The cells 18 are of a known type, for example LFP cells.

[0021] The cell 18 is characterized by, among other things, a nominal capacity Capa, an internal resistance R0, and an open circuit voltage OCV. Capa is the capacitance of the cell 18. The capacity of a cell represents the maximum amount of electrical energy that can be stored in the cell. This capacity is expressed in Ah (ampere-hours). For simplicity of explanation of this embodiment, the capacitance Capa is considered constant over time.

[0022] The internal resistance R is the value of the internal resistance of cell 18. The internal resistance of a cell is a physical quantity found in most electrical models of electrical cells. As a cell ages, its internal resistance usually increases. At instant k, the value of the internal resistance R of cell 18 is R 0,k It is expressed as:

[0023] OCV is also known as the "open circuit voltage." OCV is the voltage measurable between terminals 30 and 32 after cell 18 has been electrically isolated from an electrical load for several hours. The OCV varies according to the state of charge of the cell.

[0024] The state of charge of cell 18 at instant k is SOC k The state of charge represents the filling rate of the cell 18. When the amount of electrical energy stored in the cell 18 is equal to its capacity Capa, the state of charge is equal to 100%. When the amount of energy stored in the cell 18 is 0, i.e., when no more electrical energy can be extracted from the cell 18 to supply an electrical load, the state of charge is equal to 0%.

[0025] The initial value R of the parameter Capa and the internal resistance R 0,0 are known parameters of the cell 18. For example, they are provided by the cell manufacturer or are determined experimentally from measurements performed on this cell.

[0026] Battery 10 also includes in each cell: -A voltmeter to measure the voltage across the terminals of this cell. -An ammeter that measures the intensity of the current flowing through the cell. -A thermometer to measure the temperature inside the cell.

[0027] To simplify FIG. 1, only the voltmeter 34, ammeter 36, and thermometer 38 of the cell 18 are shown.

[0028] Here, to measure the internal temperature of the cell 18, the thermometer 38 is in direct thermal and mechanical contact with the outer casing of the cell 18. The thermometer 38 is attached directly to the cell 18.

[0029] Finally, the battery also includes a sensor 39 that measures a physical quantity representative of the ambient temperature Ta, which is the temperature of the external environment in which the cells 18 are immersed. Here, the sensor 39 is, for example, a thermometer housed between the exterior case of the battery 10 and the exterior case of each of the cells 18-21.

[0030] Unlike the various parameters of the cells 18 mentioned above, the state of charge (SOC) of the cells 18 cannot be measured directly and therefore must be estimated. For this purpose, the vehicle 2 is equipped with an electronic battery management system 40, commonly known by the abbreviation BMS (Battery Management System). One of the functions of this system 40 is to determine the state of charge of the battery 10. To determine this state of charge, the system 40 is able to estimate the state of charge of each cell of the battery 10.

[0031] To perform these various estimations, the system 40 is electrically connected to each sensor of the battery 10 to obtain the measurements necessary to estimate the state of charge of each cell.

[0032] Here, system 40 includes a memory 42 and a programmable electronic computer 44 capable of executing instructions stored in memory 42. To this end, memory 42 contains the instructions necessary to carry out the process shown in Figure 4, as well as initial values ​​for various parameters necessary to carry out the process.

[0033] 2 shows an electrical model 50 of cell 18. This model is known as the "electrical lumped parameter model" or "first-order Thévenin model." It consists of the following connected in series from terminal 32 to terminal 30: - Open circuit voltage OCV generator 52, -Parallel RC circuit 54, and -Internal resistance R0

[0034] Circuit 54 includes a capacitor of capacitance C1 connected in parallel with a resistor of value R1. In the following, these two parameters C1 and R1 of model 50 are assumed to be known and constant over time. The voltage across circuit 54 is denoted V1. The voltage between terminals 30 and 32 of cell 18 is denoted V, and the magnitude of the current through cell 18 is denoted i. The value of voltage OCV at instant k is denoted OCVk.

[0035] 3 shows a first embodiment of the configuration of estimators 60, 62, and 64 implemented in system 40 to estimate the state of charge of cell 18. Each estimator 60, 62, and 64 is implemented in the form of an estimation algorithm executed by computer 40. Hereinafter, "executing an estimator" and "executing an estimation algorithm" mean the same thing.

[0036] The estimator 60 estimates the values ​​of the parameters R0 and OCV of the electrical model 50 from measurements of the voltage V and the intensity i of the current through the cell 18. The estimator 60 is executed at each instant k1 of the time sequence of instants {0; 1; 2; ...; k1; k1+1; ...}, where these instants k1 are repeated with a constant frequency f1. The duration of the constant interval between two immediately successive instants k1 and k1+1 is denoted as Δt1. The duration Δt1 is equal to 1 / f1. The duration Δt1 is typically between 0.1 and 60 seconds, preferably between 0.1 and 10 seconds. Here, the duration Δt1 is equal to 0.2 seconds.

[0037] Below, the parameter values ​​R0 and OCV estimated at time k1 are R 0、k1 and OCV k1 The measured voltage V and intensity i at time k1 are Vm k1 and im k1 The estimator 60 is implemented as described in WO2020064959. Thus, at each k1 time point, a recursive least squares algorithm is run to estimate the measurements Vm obtained between k1 and k1-N1. k1 and im k1 From the value of OCV k1 and the coefficient value b in the following relational expression (1) 0,k1 , b 1,k1 and b 3,k1 Determine.

number

[0038] Relation (1) is derived from the electrical model 50. N1 is an integer greater than 2, preferably greater than 100 or 1000.

[0039] The coefficient value b in relational expression (1) 0,k1 ,b 1,k1 and b 3,k1 is related to the values ​​of the parameters of model 50 by the following relationships: b 0,k1 =R 0,k1 b1,k1 =-R 0,k1 +(Δt1 / C1)+(Δt1.R 0,k1 / (C 1. R1)) b 2,k1 =Δt1 / (C 1. R1)-1

[0040] Thus, at each instant k1, the estimator 60 calculates new values ​​R for the parameters R and OCV of the model 50, respectively. 0、k1 and OCV k1 Output.

[0041] The estimator 62 estimates the entropy variation ΔS of the cell 18 from measurements of the voltage V, the intensity i of the current through the cell 18, the internal temperature Ti, and the ambient temperature Ta. The estimator 62 is executed at each instant k2 in the time sequence {0; 1; 2; ...; k2; k2+1; ...}, where these instants k2 are repeated at a constant frequency f2. The duration of the constant interval between two immediately successive instants k2 and k2+1 is denoted as Δt2. The duration Δt2 is equal to 1 / f2. The temperature of the cell 18 changes more slowly than the voltage and current. Therefore, the frequency f2 is typically selected to be equal to or less than the frequency f1. For example, the duration Δt2 is equal to 5 seconds. In this case, the set of instants k2 is a subset of the set of instants k1. There are several instants k1 between two successive instants k2 and k2+1.

[0042] Hereafter, the value of the entropy change ΔS estimated at the moment k2 is ΔS k2 The measured values ​​of voltage V, intensity i, temperature Ti, and Ta at the instant k2 are V mk2 , im k2 , Tim k2 , Tam k2 Again, the estimator 62 is implemented as described in WO2020064959. Thus, the estimator 62 uses the following thermal model of the cell 18: TIFF2025542552000003.tif26164 where, -m is the mass of cell 18, -C p is the heat capacity of cell 18, -dTi / dt is the first derivative of temperature Ti with respect to time, -F is Faraday's constant, -h is the heat exchange coefficient between the cell 18 and the external environment, - A is the area of ​​the cell 18 in contact with the external environment, -Ta is the ambient temperature.

[0043] This thermal model is particularly accurate because it takes into account the heat exchange between the cell and the external environment, the generation of heat within the cell due to the Joule effect, and the entropy change due to the movement of ions such as lithium.

[0044] At each instant k2, the estimator 62 executes a recursive least squares algorithm to fit the measurements V obtained between each instant k2 and k2-N2. mk2 , imk2 , Tim k2 , Tam k2 From this, the coefficient value of the following relational expression (2) a0、k2 , a 1、k2 , and determine a3, k2.

number

[0045] The relationship (2) is derived from the thermal model described above. N2 is an integer greater than 2, preferably greater than 10, 50, or 100. In the relationship (2), the value OCV k2 is the value of the parameter OCV estimated by the estimator 60 at an instant k1 that is equal to or closest to the instant k2 and is earlier than the instant k2.

[0046] Value a 0,k2 ,a 1,k2 and a 3,k2 is related to the values ​​of the thermal model parameters by the following relationship: a 0,k2 =Δt2 / (mC p ) a 1,k2 =Δt2.ΔS k2 / (mC p .F) a 2,k2 =Δt2.hA / (mC p ).

[0047] Therefore, the estimator uses the following relationship to obtain estimates of the parameter values ​​of the thermal model: mC p = Δt2 / a 0,k2 ΔS k2 = a 1,k2 .F / a 0,k2 hA = a 2,k2 / a 0,k2 .

[0048] Thus, at each instant k2, the estimator 60 calculates a new value ΔS of the parameter ΔS of the thermal model. k2 On the other hand, the product mC p and hA generally vary slightly as a function of time. Therefore, in this embodiment, the product mC p and hA are considered constant. Their product mC p The values ​​of mC and hA are determined, for example, from data provided by the manufacturer of the cell 18 or experimentally measured during a calibration phase. p The values ​​of and hA are stored in memory 42 and are no longer estimated by estimator 62.

[0049] The estimator 64 estimates the state of charge SOC of the cell 18 from the entropy variation ΔS estimated by the estimator 62 .

[0050] The estimator 64 is executed at each instant k of the time sequence of instants {0; 1; 2; ...; k; k+1; ...}, where these instants k are repeated with a constant frequency f. The duration of the constant interval between two immediately successive instants k and k+1 is denoted as Δt. The duration Δt is equal to 1 / f. Typically, the duration Δt is between 0.2 s and 1 min. For example, here, the frequency f is equal to the frequency f1, and the duration Δt is equal to the duration Δt1. Therefore, here, the set of instants k and the set of instants k1 are identical.

[0051] Below, the value of the state of charge SOC estimated at moment k is called SOC k The measured values ​​of voltage V, intensity i, temperature Ti and Ta at the instant k are V mk , im k , Tim k , Tam k This indicates:

[0052] The estimator 64 compensates for errors introduced by the use of electrical and thermal models by the estimators 60 and 62 in order to improve the accuracy of the estimation of the state of charge SOC. To this end, the estimator 64 calculates the value of the state of charge SOC by additionally taking into account the deviations of k Estimate. -Deviation V ek -V mk where V ek is the estimate at instant k of the voltage V between terminals 30 and 32 obtained using model 50. -deviationTi ek -Tim k where Ti ek is the estimate at instant k of the temperature Ti obtained using the thermal model of the estimator 62.

[0053] To this end, the estimator 64 is implemented here as a Kalman filter. The thermal model is nonlinear. For this reason, the estimator 64 implements an extended version of a Kalman filter, commonly known by the abbreviation EKF (Extended Kalman Filter). The implementation and operation of an Extended Kalman Filter are well known to those skilled in the art. For example, the implementation and operation of an Extended Kalman Filter are described in detail in the following paper: L. Plett et al., "Extended Kalman Filter for Battery Management Systems of LiPB-Based HEV Battery Packs," Journal of Power Sources, 2004, pp. 252-292. This paper will hereinafter be referred to as "Plett2004." Therefore, only the state model and observation model of the Kalman filter of the estimator 64 will be described below.

[0054] In this example embodiment, the state vector x k [SOC k , T ik , V 1k ] T The Kalman filter uses a state representation, which gives the measurements made between times 0 and k-1 and the previous state vector x k-1 Only from the state vector x at the moment k k Prediction of x k / k-1 This state representation is constructed from the electrical and thermal models used by estimators 60 and 62. Therefore, this state representation uses the same parameters as those used by the electrical and thermal models described above. For example, here, the state representation is defined by the following relation (3):

number

[0055] The observation model used in this Kalman filter is defined by the following relation (4).

number

[0056] The operation of system 40 will be described in the specific case of estimating the state of charge of cell 18 using the process shown in FIG.

[0057] The process begins in phase 100, where values ​​for various parameters of the electrical and thermal models are initialized. For example, the parameters may be initialized based on values ​​of these parameters obtained from previous use of system 40, or from use of a system similar to system 40 with similar cells. Phase 100 begins with the calculation of the covariance matrix Q needed to run estimator 64. 64 and R 64 This also includes the initialization of the matrix Q64 represents the uncertainty of the model used. For example, the matrix Q 64 The various coefficients of are obtained from the squared error of the cell model. To this end, a useful method is to compare the actual state of charge measured in a laboratory at a given time range, t, with the charge level predicted by the model at this time, t, given the actual state of charge at the previous time, t. The root-mean-square error between the predicted and actual state of charge over a given time range provides an estimate of the model-specific error.

[0058] matrix R 64 represents the uncertainty of the measurements used. For example, the matrix R 64 The various coefficients of are obtained from the square of the standard deviation of the Gaussian noise in each measurement.

[0059] Then, the covariance matrix Q 64 and R 64 are assumed to be constant. These matrices Q 64 and R 64 is stored in advance in the memory 42.

[0060] Once the initialization phase 100 is complete, the system 40 can be used to begin a phase 102 of estimating the operational state of charge of the battery 18 in the vehicle 2 .

[0061] During the measurement phase 110, at each instant k, the voltmeter 34, ammeter 36, thermometer 38 and sensor 39 measure the voltage V, the intensity i and the temperatures Ti and Ta, respectively. These measurements Vm k , im k , Ti k and Ta k is immediately acquired by the system 40 and stored in the memory 42. Phase 110 is repeated at each instant k. Since instant k is a subset of instant k, phase 110 k2 , im k2 , Ti k2 and Ta k2It is also used to obtain

[0062] In parallel, at each instant k1, the estimator 60 calculates the parameter R0 of the electrical model 50 and the value R of the OCV. 0,k1 and OCV k1 To do this, the estimator 60 uses N1 measurements of voltage V and intensity i obtained following N1 previous iterations of the measurement phase 110. Phase 112 is performed as described in WO2020064959. Phase 112 will therefore not be described in further detail.

[0063] In parallel with phases 110 and 112, at each instant k2, the estimator 62 calculates the value ΔS of the entropy variation ΔS of the cell 18. k2 Execute a phase 114 to estimate To do this, the estimator 62 uses N2 measurements of voltage V, intensity i, and temperatures Ti and Ta taken at instants k between the current instant k2 and instant k2-N2. Phase 114 is performed as described in WO 2020064959. Therefore, a detailed description of phase 114 will be omitted.

[0064] In parallel with phases 110, 112 and 114, at each instant k, estimator 64 executes a phase 116 for estimating the state of charge of cell 18 at instant k.

[0065] To do this, in step 118, the estimator 64 calculates the state vector x using the state representation defined by relation (3): k Prediction of x k / k-1 Calculate the prediction x k / k-1 is calculated from the following: - measurement im obtained by the system 40 at the instant k k and Tam k、 the estimates Ti of the temperature Ti and the voltage V1 obtained by the estimator 64 at the end of the phase 114 at the time k-1;k-1 / k-1 and V 1,k-1 / k-1 .

[0066] In step 120, the estimator 64 also calculates a prediction P of the covariance matrix of the estimation error on the state vector xk. k / k-1 Calculate P. Typically, this is accomplished using the following relationship: k / k-1 =F k-1Pk-1 / k-1Fk-1T +Q 64 . where: -P k-1 / k-1 is the covariance matrix of the error at instant k-1 obtained by considering all measurements taken up to instant k-1 Pk-1 is an estimate of -P k / k-1 is the covariance matrix P at instant k obtained by considering only measurements taken up to instant k-1 k is the predicted value.

[0067] matrix F k-1 is the state matrix. It is obtained from the relation (3). Here, for example, the relation (3) is the vector x k Using Taylor series expansion in the vicinity of the vector x k The contribution of second and higher derivatives is ignored. Therefore, the matrix F k-1 is defined by the following relationship: TIFF2025542552000007.tif46165This matrix F k-1 In this case, the differential dΔS k / dSOC k-1 is calculated, for example, using the following relationship: dΔS k / dSOC k-1 =(ΔS k2- ΔS k2-1) / (SOC k2- SOC k2-1) , where: the instant k2 in this relationship is the most recent instant k2 at which the estimate of the entropy variation ΔS was updated by the estimator 62, -SOC k2 and SOC k2-1are the states of charge of the cell 18 determined by the estimator 64 for the instant k closest to instants k2 and k2-1, respectively.

[0068] In step 122, the estimator 64 calculates the state vector prediction x k / k-1 is corrected to obtain the corrected state vector x k / k Construct the correction vector x k / k is the deviation between Ik It is constructed as a function of - vector z of estimated values ​​of physical quantities at instant k k , and - vector z of measurements of the same physical quantity at instant k k .

[0069] deviation Ik is known as the "innovation". Here, the measured physical quantities are temperature Ti and voltage V. Therefore, the vector z k Tim k ,Vm k ] T is equal to. Innovation I k is calculated using the following relationship: Ik =z k- z k . Therefore, Innovation I k is defined by the following relationship: TIFF2025542552000008.tif3690 where the estimated value Tie k and Ve k is obtained using the observation model defined by relation (4).

[0070] Typically, in step 122, the estimator 64 calculates the innovation I k Prediction x by multiplying it by the Kalman gain Kk k / k-1 Correct the following. The gain Kk is calculated using the following relationship: k =P k / k-1 H k T (H kP k / k-1 H k T +R 64 ) -1 . where: -Matrix R 64 is the covariance matrix of the noise with respect to the measured physical quantity, -H k is the observation matrix.

[0071] Observation matrix H k is obtained from relation (4).

[0072] Next, x k / k =x k / k-1 +K k I k Using the relationship, the state vector x k / k Build.

[0073] The updated error covariance matrix at instant k is P k / k = (IK k H k )P k / k-1 It is calculated using the relation: where I is the identity matrix.

[0074] matrix P k / k is the estimated SOC k / k , T ik / k and V 1,k / k The error range of the estimated state of charge (SOC) at the moment k is k is SOC k / k is equal to.

[0075] 5 is a graph showing the change over time in the state of charge of cell 18 estimated using different algorithms. In this graph, curve 150 represents the change in the state of charge of cell 18 measured in a laboratory. This laboratory measurement is believed to be the closest to the actual state of charge value. However, the methodology used to make this estimation in the laboratory cannot be used when cell 18 is operating in a vehicle 2.

[0076] Curve 152 shows the evolution of the state of charge estimated using the process shown in Figure 4. In this process, the deviation Ve k -Vm k and deviation Tie k -Tim k I am using both.

[0077] Curve 156 shows the change in state of charge estimated using the same process as shown in FIG. 4, but with the estimated SOC k / k-1 To correct for the deviation Ve k -Vm k The difference is that only

[0078] Curve 158 shows the change in state of charge estimated using a conventional process known as coulomb counting.

[0079] Deviation Ve k -Vm k Predict SOC using k / k-1 It can be seen that correcting for σ already gives a much better estimate (curve 156) than that obtained using the conventional process (curve 158).

[0080] Also, the deviation Ve k -Vm k and deviation Tie k -Tim k Predicted SOC using both k / k-1 It can also be seen that the best estimate is obtained by correcting (curve 152).

[0081] Finally, although not shown in Figure 5 for ease of reading, the deviation Tie k -Tim k Predicted SOC using only k / k-1 When corrected, the deviation Ve k -Vm k It is emphasized that this gives a better estimate than that obtained using only, but not as good as the estimate obtained by implementing the process of FIG.

[0082] Chapter 2 Variations:

[0083] Electric model variations:

[0084] Other electrical models can be used. For example, the electrical model could alternatively include multiple parallel RC circuits connected in series between one terminal of the DC voltage source and terminal 30 of the cell. In this case, the number of electrical model parameters to be estimated would be greater. However, as previously mentioned, the values ​​of these additional parameters can be estimated by implementing a recursive least squares method.

[0085] Alternatively, the parameters R1 and C1 are not considered constant, but are instead estimated at each instant k1, for example, in a similar manner as described for the parameters R0 and OCV of the electrical model.

[0086] In a simplified embodiment, the value of the parameter R0 is considered constant over time. In this case, the value R 0,k1 is not estimated at each instant k1.

[0087] Algorithms other than the recursive least squares algorithm can be used to estimate the parameter values ​​of the electrical model. For example, the value R 0、k1 and OCV k1 can also be estimated using an additional Kalman filter dedicated to this task. 0,k1 Additional examples of Kalman filters designed to estimate the parameter OCV are described in WO2016083754 or chapter 4.2.1 of Plett2004. An example of using a Kalman filter to estimate the value of the parameter OCV is also described in US2017146608A1.

[0088] Thermal model variations:

[0089] Other thermal models are possible. For example, the thermal model used by estimator 62 may be defined by the following relationship: TIFF2025542552000009.tif28165

[0090] In another embodiment, the thermal model used to estimate the entropy variation ΔS is the thermal model of equation (5) of application US2017146608A1. In this case, the thermal model estimates the entropy value ΔS k2 The value ΔS k2-1 , and the value Ti k2、 Ti k2-1 , OCV k2 , and OCV k2-1 In this case, the ambient temperature is not used to estimate the entropy variation ΔS, and sensor 39 can be omitted.

[0091] Or, product mC p and / or the value of the product hA are not considered to be constants. In this case, the values ​​of these products are estimated at each time point k2, for example using the recursive least squares method.

[0092] The physical quantity Ta that represents the ambient temperature may differ from the temperature. For example, if the battery is equipped with a cell cooling system, the physical parameter Ta may be the control quantity of this cooling system. In fact, the greater the cooling control, the higher the ambient temperature of the cells.

[0093] Entropy value ΔS k2 can be estimated using methods other than recursive least squares. For example, ΔS k2 can be estimated using a Kalman filter dedicated to this task.

[0094] Alternatively, k2 is as frequent as k or k1.

[0095] Variations of state of charge estimation:

[0096] In a simplified variant, the deviation Ve k -Vm k Only or deviation tie k -Tim k Only use SOC to predict state of chargek / k-1 In this case, the innovation Ik There is only one deviation.

[0097] Given a matrix F k-1 In the example, the derivative dΔS k / dSOC k-1 can be calculated in different ways. For example, during the calibration phase, a polynomial approximation of the evolution of the entropy variation ΔS as a function of the state of charge SOC is constructed. Then, the derivative dΔS k / dSOC k-1 The value of SOC k-1 is equal to the value of the derivative of the polynomial constructed with the horizontal axis equal to

[0098] The estimator 64 can be implemented using forms of Kalman filters other than EKF.

[0099] In another variant, the estimator 64 is not implemented as a Kalman filter but in another form, for example, the estimator 64 may be configured to calculate, after a training phase on a database, the value SOC as a function of the measurements imk and Vmk and the deviations Vek-Vmk and Tiek-Tmk. k To do this, the database contains, for a number of instants k, the experimentally measured value SOC at instant k associated with the measurement imk, Vmk. k and the deviation Ve calculated at the same instant k using the electrical and thermal models. k -Vm k and Tie k -Tm k The estimator 64 may also be implemented using a first and a second learning machine. The first learning machine learns the measurement im k , v mk , Tim k and Tam k Deviation from Ve k -Vm k and Tie k -Tm k and the value of the entropy fluctuation ΔS estimated at the moment k2 closest to the moment k, ΔS k2Next, the second learning machine is configured to generate the deviation Ve generated by the first learning machine. k -Vm k and Tie k -Tm k and the estimated value ΔS k2 From this, the estimated state of charge of cell 18, SOC k is configured to provide:

[0100] Alternatively, moment k is less frequent than moment k1.

[0101] The state-of-charge estimation process described herein for the particular case of a single battery cell can also be applied to a battery including a pack of cells, where the battery is treated as if it were a single cell. In other words, what is described herein applies to the case of the battery itself being composed of multiple cells electrically connected together.

[0102] Other variations:

[0103] The thermometer 38 may be housed within the casing of the cell 18 .

[0104] The sensor 39 may be located outside the outer casing of the battery 10 .

[0105] What has been described in the specific case of LFP cells applies to any cell technology. For example, it applies to LiPB (Lithium Ion Polymer Batteries) or Li-IP cells or the like.

[0106] Alternatively, the cell's capacitance (Capa) may not be constant but may vary over time. In this case, the value of the capacitance (Capa) may be estimated. An example of a process for estimating the capacitance (Capa) is described in Chapter 4.2.2 of Part 3 of Plett 2004. Another example is described in WO 2016083754.

[0107] The instructions given here in the specific case of electric vehicle cells and batteries apply to all cells and batteries, whether or not they are used in electric vehicles. This applies not only to new cells and new batteries, but also to used cells and used batteries.

[0108] Some of the above variations may be combined in a single embodiment.

[0109] Section 3: Advantages of the described embodiments:

[0110] Estimating the state of charge SOC of the cell from the entropy change ΔS improves the accuracy of the state of charge estimation, especially when the voltage OCV does not vary significantly with the state of charge SOC. Furthermore, the deviation Ve k -Vm k and Tie k -Tim k The value SOC as a function of at least one of k Calculating SOC corrects for errors caused by the use of electrical and thermal models by estimators 60 and 62. As a result, the value SOC calculated in this manner k is particularly more accurate compared to the process proposed in, for example, WO2020064959.

[0111] Deviation Ve k -Vm k and deviation Tie k -Tim k Estimated SOC using both k Correcting β increases the accuracy of the estimated state of charge.

[0112] Using a thermal model that takes into account measurements of the intensity of the current i flowing through the cell and the ambient temperature improves the accuracy of the estimation of the entropy variation ΔS and therefore the accuracy of the estimated state of charge.

[0113] SOC k Using a Kalman filter to estimate the SOC valuek Speed ​​up the estimation process.

Claims

1. 1. An automated process for estimating the state of charge of a battery cell by an electronic battery management system, comprising: The automated process comprises: a first phase (112) of estimating the internal resistance and the open-circuit voltage of said cell using an electrical model linking the voltage V between the terminals of said cell and the intensity i of the current passing through said cell, said electrical model comprising a parameter R corresponding to said internal resistance of said cell; ０ and a parameter OCV corresponding to the open circuit voltage of the cell; a second phase (114) of estimating the variation in entropy ΔS of the cell using a thermal model of the cell, the thermal model relating the variation in the internal temperature of the cell to the variation in entropy ΔS, the thermal model relating the parameter R estimated during the first estimation phase ０ and OCV; - successive moments in time {0; 1; 2; ... ; k; k+1; . . .}, a third phase (116) of estimating the state of charge of the cell at the moment k from the change in entropy ΔS estimated during the second estimation phase, wherein the third estimation phase (116) estimates, for at least one physical quantity selected from the group consisting of the internal temperature of the cell and the voltage between the cell terminals, calculating (118) an estimate of the physical parameter using the electrical model if the physical parameter is the voltage across the cell terminals, or using the thermal model if the physical parameter is the internal temperature, then calculating (122) the deviation between the estimated value of the physical parameter and the measured value of the physical parameter, then - using the calculated deviation to construct (122) an estimate of the state of charge at instant k, Automatic process.

2. The third estimation phase comprises: Using the electrical model, an estimate of the voltage V between the cell terminals at an instant k is calculated. ｋ Then, the estimated value Ve is calculated (118). ｋ and the measured value Vm ｋ Calculating (122) a first deviation between - an estimate of the internal temperature of the cell at instant k using a thermal model, Tie ｋ Calculate (118) the estimated value Tie ｋ and the measured value Tim ｋ and calculating (122) a second deviation between the parameter R ０ constructing (122) an estimate of the state of charge using the first and second deviations calculated to correct for errors introduced by the use of the electrical model and the thermal model in estimating the OCV and entropy variation ΔS; 2. The automated process of claim 1, comprising the steps of:

3. The thermal model calculates the change in the internal temperature of the cell by: the intensity of the current passing through the cell, and - relating it to the ambient temperature of the medium in which said cell is immersed, 3. An automated process according to claim 1 or 2.

4. The state of charge estimate is obtained by implementing a Kalman filter including a first and a second deviation.

4. An automated process according to any one of claims 1 to 3.

5. State variable state x k / k+1 The state representation used by the Kalman filter to predict 5. The automated process of claim 4. -x ｋ／ｋ－１ is the estimate of the state vector obtained at instant k, made by considering only measurements made between instants 0 and k−1, -SOC ｋ／ｋ－１ is the predicted state of charge obtained at instant k, made by considering only measurements made between instants 0 and k−1, -SOC ｋ－１／ｋ－１ is the state of charge estimate obtained at instant k−1, made by considering only all measurements made between instants 0 and k−1, -Ti ｋ／ｋ－１ is the predicted value of the internal temperature obtained at instant k, made by considering only measurements made between instants 0 and k-1, -Ti ｋ－１／ｋ－１ is an estimate of the internal temperature obtained at instant k−1, made by considering only all measurements made between instants 0 and k−1; -V １，ｋ／ｋ－１ is the voltage V at the terminals of the RC circuit of the electrical model obtained at the instant k １ is a prediction of , made by considering only all measurements made between instants 0 and k-1, -V １，ｋ－１／ｋ－１ is the voltage V obtained at the moment k-1 １ is an estimate of , made by considering all measurements made between instants 0 and k-1, -Δt is the time elapsed between two successive instants k and k−1, -Capa is the value of the cell capacity, -im ｋ is a measure of the intensity of the current passing through the cell at the instant k, -R ０，ｋ is the internal resistance R obtained at the end of the first estimation phase carried out at a moment k1 equal to the moment k or at a moment k immediately before the moment k; ０ is an estimate of -m is the mass of the cell; -C ｐ is the heat capacity of the cell, -h is the heat exchange coefficient between the cell and the external medium, - A is the area of ​​the cell in contact with the external medium, -Tam ｋ is the ambient temperature of the external environment at the instant k, -ΔS ｋ is the value of the entropy variation ΔS obtained at the end of said second estimation phase, carried out at an instant k2 equal to or shortly before the instant k, -F is the Faraday constant.

6. The first estimation phase comprises: - successive moments in time {0; 1; 2; ...; k1; k1+1 ; . . .}, at each instant k1, the measured value Vm of the voltage between the cell terminals ｋ１ and a measurement of the intensity of the current flowing through the cell, i ｋ１ , and the index k1 is the measured value Vm ｋ１ and im ｋ１ Identify the instant k1 at which -N １ is a predetermined integer greater than 3, and the instants k1 and k1-N １ the parameters R of the electrical model of the cell by performing a recursive least squares algorithm from voltage and intensity measurements taken between ０ and the OCV value R ０ , k1 and OCV ｋ１ are estimated, respectively.

6. An automated process according to any one of claims 1 to 5.

7. The parameter R of the electrical model ０ and the OCV value R ０、ｋ１ and OCV ｋ１ is estimated using the following relation:

7. The automated process of claim 6. where: -Vm ｋ１ and Vm ｋ１－１ are measurements of the voltages across the cell terminals at instants k1 and k1-1, respectively; -im ｋ１ and im ｋ１－１ are measurements of the intensity of the current flowing through the cell at the instants k1 and k1-1, respectively, -OCV ｋ１－１ is the value of the OCV parameter estimated at instant k1-1, -b ０，ｋ１ , b １，ｋ１ and b ２，ｋ１ is the value of the electrical model coefficient updated at the instant k1, and the parameter R ０ The value of R ０，ｋ１ is the value b ０，ｋ１ is equal to.

8. The second estimation phase comprises: - successive moments in time {0; 1; 2; ...; k2; k2+1 ; . . .}, at each instant k2, the measured value Vim of the voltage between the cell terminals ｋ２ , a measurement of the intensity of the current flowing through the cell i m ｋ２ , the measurement of the internal temperature of the cell Tim ｋ２ , and a measurement of the physical quantity Tam representing the ambient temperature of the medium in which the cell is immersed ｋ２ Obtain the measurement Vi ｍｋ２ , im ｋ２ , Tim ｋ２ , and Tam ｋ２ obtain an index k2 that identifies the instant k2 at which -N2 is a predetermined integer greater than 2, and the instants k2 and k2-N ２ A value ΔS of the entropy variation ΔS is calculated by executing a recursive least squares algorithm from the measured values ​​of physical quantities representing the voltage between the cell terminals, the intensity of the current flowing through the cell, the internal temperature, and the ambient temperature, which are obtained between the cell terminals and the value of the internal resistance or the value of the open-circuit voltage obtained after the first estimation phase. ｋ２ Estimate the 8. An automated process according to any one of claims 1 to 7.

9. The value of the entropy change ΔS is ΔS ｋ２ is estimated using the following relation:

9. The automated process of claim 8. where: -Tim ｋ２ and Tim ｋ２－１ are the values ​​of the internal cell temperatures measured at k2 and k2-1, respectively; -Im ｋ２ is the value of the intensity of the current flowing through the cell measured at k2, -Tam ｋ２ is the value of the physical quantity representing the ambient temperature measured at k2, -OCV ｋ２ is the value of the parameter OCV obtained after the execution of the first estimation phase at an instant equal to or immediately preceding k2, -a ０，ｋ２ , ａ１，ｋ２ and a ２，ｋ２ is the value of the thermal model coefficient updated at k2, and the value of the entropy change ΔS at k2 is ΔS ｋ２ teeth, ａ１，ｋ１． F / a ０，ｋ２ where F is Faraday's constant.

10. An information recording medium (42) readable by an electronic computer, which, when executed by said electronic computer, 10. A method for estimating a signal, comprising: Information recording medium.

11. 1. An electronic system for managing a battery comprising at least one cell, comprising: the electronic system comprises an electronic computer (44) programmed to execute an automatic process for estimating the state of charge of the battery cells; The electronic computer (44) is programmed to carry out an automatic process for estimating the state of charge according to any one of claims 1 to 9. Electronic systems.

12. A motor vehicle, - at least one driving wheel (6), an electric motor (4) capable of rotating said drive wheels and moving said vehicle; a battery (10) including at least one cell (18-21) capable of storing electrical energy and alternately supplying said electrical energy to said electric motor, said cell including two terminals (30, 32) electrically connected to said electric motor; a voltmeter (34) electrically connected between the cell terminals to measure the voltage between the cell terminals; an ammeter (36) connected in series with the electric cell and measuring the intensity of the current passing through said cell; - a thermometer (38) for measuring the internal temperature of the electric cell, and a battery management system (40) connected to a voltmeter and an ammeter, the battery management system includes a programmable electronic computer (44) capable of estimating the state of charge of the electric cells from the measurements of the voltmeter and the ammeter; The battery management system (40) is characterized in that it complies with claim 11. car.

Images