Method and System for Estimating State of Charge of Battery Based on Battery Parameter Compensation

Abstract

The present invention relates to the technical field of battery management, and particularly to a method and system for estimating the state of charge of a battery based on battery parameter compensation. An extended Kalman filter model is established through the maximum available charge of the battery, the OCV-SOC relationship curve at different temperatures, and the resistance-capacitance parameter relationship curve of the battery model. The parameters in the extended Kalman filter model matrix are updated by each change in battery temperature, and the extended Kalman filter model is corrected in real time. The state of charge of the battery is estimated through the corrected extended Kalman filter model, so that the estimation of the state of charge of the battery is more accurate.

Description

Technical Field

[0001] The present invention relates to the technical field of battery management, and particularly to a method and system for estimating the state of charge of a battery based on battery parameter compensation. Background Art

[0002] With the development of technology, electric vehicles have gradually become the mainstream in the automotive market. In order to manufacture high-performance and highly reliable electric vehicles, it is necessary to monitor and track the on-vehicle battery pack in real time. Accurately predicting the state of charge (SOC) of the battery and the battery health (SOH) is of great significance for improving the overall vehicle performance and enhancing vehicle safety.

[0003] Existing methods for estimating the state of charge (SOC) of a battery include: model-driven methods, data-driven methods, etc., among which the equivalent circuit model is the most widely used. A series of methods for estimating the state of charge of a battery using the Kalman filter algorithm have emerged. However, due to the non-linear dynamic characteristics of the battery under different working conditions, and the influence of factors such as environmental changes and battery aging on the battery model, the existing methods have large estimation errors for the state of charge of the battery and cannot meet the usage requirements. Summary of the Invention

[0004] The purpose of the present invention is to provide a method and system for estimating the state of charge of a battery based on battery parameter compensation, so as to solve the problem of large estimation errors in the existing state of charge estimation of the battery.

[0005] The technical problem solution of the present invention:

[0006] A method for estimating the state of charge of a battery based on battery parameter compensation, characterized by including the following steps:

[0007] S1. Based on the second-order RC equivalent circuit model of the battery to be evaluated, establish a data set of battery parameters to be evaluated, and the data set of battery parameters to be evaluated includes the OCV-SOC relationship curve at different temperatures and the resistance-capacitance parameter relationship curve of the battery model obtained by performing HPPC tests at different temperatures and different states of charge;

[0008] S2. Monitor the battery charge and discharge data to update the maximum available charge Q of the battery max ;

[0009] S3. Based on the second-order RC equivalent circuit model of the battery, use the data set of battery parameters and the maximum available charge Q max to establish a battery state space model;

[0010] S4. Determine the extended Kalman filter model of the battery based on the battery state space model; according to the real-time temperature of the battery and the maximum available charge Q of the battery maxUpdate the matrix parameters in the battery extended Kalman filter model to correct the extended Kalman filter model, and estimate the state of charge of the battery according to the corrected extended Kalman filter model.

[0011] Further defined, the method for obtaining the OCV-SOC relationship curve at different temperatures is as follows:

[0012] After obtaining the OCV-SOC data through intermittent charge and discharge experiments on the battery at a set of multiple temperatures, fit the data to obtain the expression of the OCV-SOC relationship curve at the corresponding temperature. Through this expression, the open circuit voltage value of the current battery can be calculated using the state of charge of the battery. The expression of the OCV-SOC relationship curve is:

[0013] OCV(t,SOC)=p1 t *SOC^7+p2 t *SOC^6+p3 t *SOC^5+p4 t *SOC^4+p5 t *SOC^3+p6 t *SOC^2+p7 t *SOC+p8 t

[0014] Where, t is the set temperature, OCV is the open circuit voltage of the battery, SOC is the state of charge of the battery, the value range of SOC is 0 to 100%, and p1 t ~p8 t are fitting coefficients;

[0015] The method for obtaining the resistance-capacitance parameter relationship curve of the battery model obtained by performing HPPC tests at different states of charge is as follows:

[0016] Perform HPPC tests on the battery at different states of charge at a set temperature, obtain the corresponding battery response terminal voltage curve at this temperature and process it to obtain the corresponding resistance-capacitance parameters, and fit the resistance-capacitance parameters to obtain the corresponding resistance-capacitance parameter relationship curve.

[0017] Further defined, the method for monitoring the battery charge and discharge data to update the maximum available charge Q max of the battery is as follows:

[0018] According to Calculate, where Q passed is the charge flowing between the state of charge SOC 1 and the state of charge SOC 2 of the battery, and SOC 1 and SOC 2The SOC is the state of charge of the battery at two different times when the battery is charged or discharged. 1 With SOC 2 The battery voltage change needs to meet Where dt is the time difference between two different moments, and dV is the voltage difference within the corresponding time difference.

[0019] It is further defined that the battery space state model includes a state equation and an observation equation, and the state equation is:

[0020]

[0021] The observation equation is:

[0022] U(K)=OCV(T,SOC)-U s (T,K)-U d (T,K)-R 0 (T,SOC)I(K)+ν K

[0023] Among them, U s is the concentration polarization voltage of the battery, U d is the electrochemical polarization voltage of the battery, K is the Kth calculation of the battery state of charge, τ s (T,SOC)=R s (T,SOC)*C s (T,SOC), τ d (T,SOC)=R d (T,SOC)*C d (T, SOC), I(K) is the battery current at the Kth cycle, Δt is the sampling time interval, η is the Coulomb efficiency, ω K is the process noise, ν K is the observation noise, e is a constant, R s is the concentration polarization resistance of the battery, R d is the electrochemical polarization resistance of the battery, C s is the concentration polarization capacitance of the battery, C d is the electrochemical polarization capacitance of the battery.

[0024] It is further defined that the battery extended Kalman filter model includes a state transfer matrix, an input matrix and an input matrix:

[0025] The state transfer matrix is:

[0026]

[0027] The input matrix is:

[0028]

[0029] The observation matrix is as follows:

[0030]

[0031] After the battery temperature T changes, the parameters of the matrix in the extended Kalman filter model are updated. The steps for the extended Kalman filter model to predict the state of charge of the battery and update the parameters in the matrix are as follows:

[0032] S4.1. Calculate the estimated state variable X K =(SOC(T, K + 1)U s (T, K + 1)U d (T, K + 1)) T ,

[0033] S4.2. Calculate the observed output value

[0034] S4.3. Update the parameters to solve for the new state transition matrix A K-1 ,

[0035] S4.4. Calculate the pre-estimated value of the error covariance

[0036] S4.5. Update the parameters to solve for the new observation matrix C K ,

[0037] S4.6. Calculate the Kalman gain L K ,

[0038] S4.7. Calculate the optimal estimate of the state variable

[0039] S4.8. Calculate the optimal estimate of the error covariance

[0040] Where X 0 =(0.8 0.01 0.01) T , X 0 and are the initial values of the algorithm iteration respectively, R K is the observation noise error covariance, y K is the battery terminal voltage measured at the Kth step, is the voltage prediction value of the current battery, Q K is the process noise.

[0041] A state of charge estimation system for a battery based on battery parameter compensation, characterized by comprising:

[0042] A battery parameter set storage module, configured to establish a data set of battery parameters to be evaluated according to the second-order RC equivalent circuit model of the battery to be evaluated, and the data set of battery parameters to be evaluated includes the OCV-SOC relationship curve at different temperatures and the resistance-capacitance parameter relationship curve of the battery model obtained by performing HPPC tests at different temperatures and different states of charge;

[0043] A battery maximum available charge amount update module, configured to monitor the battery charge and discharge data to update the battery maximum available charge amount Q max ;

[0044] A battery space state model establishment module, configured to establish a battery space state model according to the second-order RC equivalent circuit model of the battery, using the data set of battery parameters and the maximum available charge amount Q max to establish a battery space state model;

[0045] A Kalman filter model establishment module, configured to determine a battery extended Kalman filter model according to the battery space state model; update the matrix parameters in the battery extended Kalman filter model according to the real-time temperature of the battery and the battery maximum available charge amount Q max to realize the correction of the extended Kalman filter model, and estimate the state of charge of the battery according to the corrected extended Kalman filter model.

[0046] Further defined, the battery parameter set storage module includes:

[0047] An OCV-SOC relationship curve storage module, configured to obtain OCV-SOC data by performing intermittent charge and discharge experiments on the battery at a set plurality of temperatures, fit the data to obtain an expression of the OCV-SOC relationship curve at the corresponding temperature, store the expression of the OCV-SOC relationship curve at the corresponding temperature obtained by fitting the data, and through this expression, the open-circuit voltage value of the current battery can be calculated through the state of charge of the battery;

[0048] The expression of the OCV-SOC relationship curve is: OCV(t,SOC) = p1 t *SOC^7 + p2 t *SOC^6 + p3 t *SOC^5 + p4 t *SOC^4 + p5 t *SOC^3 + p6 t *SOC^2 + p7 t *SOC + p8t

[0049] Among them, t is the set temperature, OCV is the open-circuit voltage of the battery, SOC is the state of charge of the battery, and the value range of SOC is 0 to 100%, p1 t ~p8 t are fitting coefficients;

[0050] The resistance-capacitance parameter relationship curve storage module is used to perform HPPC tests on the battery under different states of charge at a set temperature, obtain the corresponding battery response terminal voltage curve at this temperature and process it to obtain the corresponding resistance-capacitance parameters, and obtain the corresponding resistance-capacitance parameter relationship curve by fitting the resistance-capacitance parameters and store it.

[0051] Further defined, the battery maximum available charge amount update module is specifically:

[0052] According to Calculate, where Q passed is the state of charge SOC of the battery 1 to the state of charge SOC of the battery 2 The amount of charge flowing between them, SOC 1 and SOC 2 are the states of charge of the battery at two different times when charging or discharging the battery, and SOC 1 and SOC 2 The battery voltage change amount needs to satisfy where dt is the time difference between two different times, and dV is the voltage difference within the corresponding time difference.

[0053] Further defined, the battery spatial state model establishment module includes:

[0054] State equation establishment module: used to establish a state equation, and the state equation is:

[0055]

[0056] Observation equation establishment module, used to establish an observation equation, and the observation equation is:

[0057] U(K) = OCV(T, SOC) - U s (T, K) - U d (T, K) - R 0 (T, SOC)I(K) + ν K

[0058] Among them, U s is the concentration polarization voltage of the battery, U d is the electrochemical polarization voltage of the battery, K is the state of charge of the battery calculated for the Kth time, τ s (T, SOC) = R s (T, SOC) * C s (T, SOC), τ d (T, SOC) = R d (T, SOC) * C d (T, SOC), I(K) is the battery current at the K - th cycle, Δt is the sampling time interval, η is the Coulomb efficiency, ω K is the process noise, ν K is the observation noise, e is a constant, R s is the concentration polarization resistance of the battery, R d is the electrochemical polarization resistance of the battery, C s is the concentration polarization capacitance of the battery, C d is the electrochemical polarization capacitance of the battery.

[0059] Further defined, the extended Kalman filter model establishment module includes:

[0060] A state transition matrix establishment module, used to establish a state transition matrix, and the state transition matrix is:

[0061]

[0062] An input matrix establishment module, used to establish an input matrix, and the input matrix is:

[0063]

[0064] An observation matrix establishment module, used to establish an observation matrix, and the observation matrix is:

[0065]

[0066] An update module, used to update the parameters of the matrix in the extended Kalman filter model after the battery temperature T changes. The steps for the extended Kalman filter model to predict the state of charge of the battery and update the parameters in the matrix are as follows:

[0067] S4.1. Calculate the estimated state variable X K =(SOC(T, K + 1)U s (T, K + 1)U d (T, K + 1)) T ,

[0068] S4.2. Calculate the observed output value

[0069] S4.3. Update the parameters to solve the new state transition matrix AK-1 ,

[0070] S4.4. Calculate the pre - estimated value of the error covariance

[0071] S4.5. Update the parameters to obtain the new observation matrix C K ,

[0072] S4.6. Calculate the Kalman gain L K ,

[0073] S4.7. Calculate the optimal estimate of the state variable

[0074] S4.8. Calculate the optimal estimate of the error covariance

[0075] Among them, X 0 =(0.8 0.01 0.01) T , X 0 and are respectively the initial values of the algorithm iteration, R K is the observation noise error covariance, y K is the battery terminal voltage measured at the K - th step, is the voltage pre - estimated value of the current battery, Q K is the process noise.

[0076] The beneficial effects of the present invention are as follows:

[0077] By using the maximum available charge of the battery, the OCV - SOC relationship curve at different temperatures and the relationship curve between the resistance - capacitance parameters of the battery model, an extended Kalman filter model is established. The parameters in the extended Kalman filter model matrix are updated with each change in battery temperature, and the extended Kalman filter model is corrected in real - time. The state of charge of the battery is estimated through the corrected extended Kalman filter model, making the estimation of the state of charge of the battery more accurate. BRIEF DESCRIPTION OF THE DRAWINGS

[0078] Figure 1 It is the second - order RC equivalent circuit model diagram of the battery to be evaluated in Embodiment 1 of the present invention;

[0079] Figure 2 It is the OCV - SOC relationship curve corresponding to T=-10°C, -5°C, 0°C, 5°C, 15°C, 25°C, 35°C, 45°C in Embodiment 1 of the present invention;

[0080] Figure 3 This is the HPPC cycle test diagram in Embodiment 1 of the present invention, where Figure 3 a is the curve diagram of the voltage change at the battery response terminal, Figure 3 b is the curve diagram of the battery current change;

[0081] Figure 4 This is the curve of the resistance-capacitance parameter relationship at different temperatures in Embodiment 1 of the present invention, Figure 4 a is the curve of the resistance value parameter relationship, Figure 4 b is the curve of the capacitance value parameter relationship;

[0082] Figure 5 This is the comparison diagram of the estimated results of the state of charge of the battery based on the battery parameter compensation method and the AH integration algorithm in Embodiment 1 of the present invention;

[0083] Figure 6 is Figure 5 the enlarged schematic diagram of part A in

[0084] Figure 7 This is the comparison diagram of the estimation error between the state of charge estimation method based on battery parameter compensation and the AH integration algorithm in Embodiment 1 of the present invention. Specific implementation manner

[0085] Embodiment 1

[0086] This embodiment provides a method for estimating the state of charge of a battery based on battery parameter compensation, including the following steps:

[0087] S1. Based on the second-order RC equivalent circuit model of the battery to be evaluated, establish a data set of battery parameters to be evaluated, where the data set of battery parameters to be evaluated includes the OCV-SOC relationship curve at different temperatures and the resistance-capacitance parameter relationship curve of the battery model obtained by HPPC testing at different temperatures and different states of charge;

[0088] S2. Monitor the battery charge and discharge data to update the maximum available charge amount Q of the battery max ;

[0089] S3. Based on the second-order RC equivalent circuit model of the battery, use the battery parameter data set and the maximum available charge amount Q max to establish a battery state space model;

[0090] S4. Determine the extended Kalman filter model of the battery based on the battery state space model; update the matrix parameters in the extended Kalman filter model of the battery according to the real-time temperature of the battery and the maximum available charge amount Q max to realize the correction of the extended Kalman filter model, and estimate the state of charge of the battery according to the corrected extended Kalman filter model.

[0091] Specifically, in step S1, a lithium-ion battery is taken as an example for specific illustration. As Figure 1 shown, a second-order RC equivalent circuit model of the lithium-ion battery is established, where U s is the concentration polarization voltage of the battery, U d is the electrochemical polarization voltage of the battery, R s is the concentration polarization resistance of the battery, R d is the electrochemical polarization resistance of the battery, C s is the concentration polarization capacitance of the battery, C d is the electrochemical polarization capacitance of the battery, U oc is the open-circuit voltage of the battery, R 0 is the ohmic resistance of the battery, U t is the port voltage of the battery that can be directly measured.

[0092] A dataset of battery parameters to be evaluated is established. The dataset of battery parameters to be evaluated includes the OCV-SOC relationship curves at different temperatures and the resistance-capacitance parameter relationship curves of the battery model obtained from HPPC tests at different temperatures and different states of charge. Among them, the method for obtaining the OCV-SOC relationship curves at different temperatures is as follows:

[0093] After obtaining the OCV-SOC data through intermittent charge and discharge experiments on the battery at a set of multiple temperatures, the data is fitted to obtain the expression of the OCV-SOC relationship curve at the corresponding temperature. Through this expression, the open-circuit voltage value of the current battery can be calculated using the state of charge of the battery. The expression of the OCV-SOC relationship curve is:

[0094] OCV(t,SOC) = p1 t *SOC^7 + p2 t *SOC^6 + p3 t *SOC^5 + p4 t *SOC^4 + p5 t *SOC^3 + p6 t *SOC^2 + p7 t *SOC + p8 t ;

[0095] Among them, t is the set temperature, usually taken according to the actual use temperature. For example, the values are -10°C, -5°C, 0°C, 5°C, 15°C, 25°C, 35°C, and 45°C respectively. OCV is the open-circuit voltage of the battery at the corresponding temperature, SOC is the state of charge of the battery at the corresponding temperature, the value range of SOC is 0 to 100%, and p1 t ~p8 t are the fitting coefficients obtained from fitting. Referring to Table 1, the OCV-SOC data when t = 25°C;

[0096] Table 1 OCV-SOC data at t = 25°C

[0097]

[0098] The corresponding OCV-SOC relationship data at t = 25°C obtained from Table 1 can be fitted for the expression through the fitting toolbox in Matlab to obtain the OCV-SOC relationship curve expression at t = 25°C: OCV(t = 25, SOC) = p1 t＝25 *SOC^7 + p2 t＝25 *SOC^6 + p3 t＝25 *SOC^5 + p4 t＝25 *SOC^4 + p5 t＝25 *SOC^3 + p6 t＝25 *SOC^2 + p7 t＝25 *SOC + p8 t＝25 where, according to the calculation, p1 t＝25 = 47.44, p2 t＝25 = -166.3, p3 t＝25 = 234, p4 t＝25 = -172, p5 t＝25 = 73.43, p6 t＝25 = -18.83, p7 t＝25 = 3.208, p8 t＝25 = 3.261; similarly, the OCV-SOC relationship data at other temperatures are measured and fitted to obtain the OCV-SOC relationship curve expression at the corresponding temperature, and finally the OCV-SOC relationship curve is obtained according to the corresponding OCV-SOC relationship curve expression, referring to Figure 2 , which are the OCV-SOC relationship curves corresponding to t = -10°C, -5°C, 0°C, 5°C, 15°C, 25°C, 35°C, and 45°C respectively.

[0099] Among them, the method for obtaining the resistance-capacitance parameter relationship curve of the battery model obtained by HPPC testing under different state of charge is as follows:

[0100] At the set temperature, the battery is subjected to HPPC testing under different state of charge, and the corresponding battery response terminal voltage curve at this temperature is obtained and processed to obtain the corresponding resistance-capacitance parameters, and the corresponding resistance-capacitance parameter relationship curve is obtained by fitting the resistance-capacitance parameters.

[0101] For example, at the set temperature, the lithium-ion battery is subjected to HPPC cycle testing at 0%, 5%, 10%, 15%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, 95%, and 100% state of charge, referring to Figure 3 ,Figure 3 Curve a shows the battery terminal voltage response curve corresponding to the HPPC cycle test of a lithium-ion battery at t = 25°C. Figure 3 Curve b shows the battery current change curve. According to the HPPC cycle test chart, the corresponding battery response terminal voltage data of the lithium-ion battery at different states of charge at this temperature are obtained, and the corresponding R 0 , R s and R d resistance values are calculated to obtain the capacitance values of C s and C d . According to the calculated resistance and capacitance data, the curve expressions of R 0 (t, SOC), R s (t, SOC), C s (t, SOC), R d (t, SOC) and C d (t, SOC) at different temperatures are obtained. Refer to Figure 4 . Figure 4 Curve a shows the curve of the resistance parameter relationship of R 0 , R s and R d at t = 25°C. Figure 4 Curve b shows the curve of the capacitance parameter relationship of C s and C d at t = 25°C.

[0102] Specifically, in step S2, the method of monitoring the battery charge and discharge data to update the maximum available charge amount Q max of the battery is as follows:

[0103] It is calculated according to , where Q passed is the charge amount flowing between the state of charge SOC 1 and the state of charge SOC 2 of the battery. SOC 1 and SOC 2 are the states of charge of the battery at two different times when the battery is charged or discharged, and SOC 1 and SOC 2 need the battery voltage change amount to satisfy where dt is the time difference between two different times, and dV is the voltage difference within the corresponding time difference.

[0104] Specifically, in step S3, after discretizing the equivalent circuit model, a lithium battery space state model composed of a state equation and an observation equation is obtained. The battery space state model includes a state equation and an observation equation. The state equation is:

[0105]

[0106] The observation equation is:

[0107] U(K) = OCV(T, SOC) - U s (T, K) - U d (T, K) - R 0 (T, SOC)I(K) + ν K

[0108] Wherein, U s is the concentration polarization voltage of the battery, U d is the electrochemical polarization voltage of the battery, K is the state of charge of the battery calculated for the Kth time, τ s (T, SOC) = R s (T, SOC) * C s (T, SOC), τ d (T, SOC) = R d (T, SOC) * C d (T, SOC), I(K) is the battery current during the Kth cycle, Δt is the sampling time interval, η is the Coulomb efficiency, ω K is the process noise, ν K is the observation noise, e is a constant, R s is the concentration polarization resistance of the battery, R d is the electrochemical polarization resistance of the battery, C s is the concentration polarization capacitance of the battery, C d is the electrochemical polarization capacitance of the battery; It can be seen that the battery parameters in the system are compensated according to the changes in temperature and the state of charge of the battery. The established second-order RC equivalent circuit model of the battery to be evaluated can not only ensure the accuracy but also reduce the computational complexity of the higher-order RC equivalent model.

[0109] Specifically, in step S4, after linearizing the battery state space equation, the extended Kalman filter model can be obtained. The battery extended Kalman filter model includes a state transition matrix, an input matrix, and an input matrix:

[0110] The state transition matrix is:

[0111]

[0112] The input matrix is:

[0113]

[0114] The observation matrix is:

[0115]

[0116] After the battery temperature T changes, the parameters of the matrix in the extended Kalman filter model are updated. The steps for the extended Kalman filter model to predict the state of charge of the battery and update the parameters in the matrix are as follows:

[0117] S4.1. Calculate the predicted state variable X K =(SOC(T, K + 1)U s (T, K + 1)U d (T, K + 1)) T ,

[0118] S4.2. Calculate the observed output value

[0119] S4.3. Update the parameters to solve for the new state transition matrix A K-1 ,

[0120] S4.4. Calculate the predicted value of the error covariance where is the transpose of the A K-1 matrix;

[0121] S4.5. Update the parameters to solve for the new observation matrix C K ,

[0122] S4.6. Calculate the Kalman gain L K ,

[0123] S4.7. Calculate the optimal estimate of the state variable

[0124] S4.8. Calculate the optimal estimate of the error covariance

[0125] where, X 0 =(0.8 0.01 0.01) T , X 0 and are the initial values of the algorithm iteration, R K is the observation noise error covariance, y K is the battery terminal voltage measured at the Kth step, is the voltage prediction value of the current battery, Q K is the process noise.

[0126] Loop through steps S3.1 to S3.8 every time the temperature and the state of charge of the battery change, so as to update the extended Kalman filter model, and then estimate the current state of charge of the battery through the updated Kalman filter model to obtain a more accurate estimated value. Refer to Figure 5 Figure 5 is a comparison chart of the estimated results of the state of charge of the battery based on the battery parameter compensation method of the present invention and the AH integration algorithm. Among them, EKF is the curve of the state of charge of the battery estimated by the battery state of charge estimation method based on battery parameter compensation of the present invention, Real is the curve of the actual state of charge of the battery, AH is the curve of the state of charge of the battery estimated by the AH integration algorithm, and the enlarged part is the comparison between the curve of the state of charge of the battery estimated by the battery state of charge estimation method based on battery parameter compensation of the present invention and the curve of the actual state of charge of the battery. Refer to Figure 6 and Figure 7 Figure 6 is a comparison chart of the error between the state of charge of the battery obtained by the battery state of charge estimation method based on battery parameter compensation of the present invention and the actual state of charge of the battery and the error between the state of charge of the battery estimated by the AH integration algorithm and the actual state of charge of the battery. The error between the state of charge of the battery obtained by the battery state of charge estimation method based on battery parameter compensation and the actual state of charge of the battery is less than 3%, meeting the usage requirements, and the estimated result is more accurate.

[0127] Embodiment 2

[0128] This embodiment provides a battery state of charge estimation system based on battery parameter compensation, including:

[0129] A battery parameter set storage module, configured to establish a data set of battery parameters to be evaluated according to the second-order RC equivalent circuit model of the battery to be evaluated. The data set of battery parameters to be evaluated includes the OCV-SOC relationship curve at different temperatures and the resistance-capacitance parameter relationship curve of the battery model obtained by performing HPPC tests at different temperatures and different states of charge;

[0130] A battery maximum available charge amount update module, configured to monitor the battery charge and discharge data to update the battery maximum available charge amount Q max ;

[0131] A battery state space model establishment module, configured to establish a battery state space model according to the second-order RC equivalent circuit model of the battery, using the battery parameter data set and the maximum available charge amount Q max to establish a battery state space model;

[0132] A Kalman filter model establishment module, configured to determine the battery extended Kalman filter model according to the battery state space model; according to the real-time temperature of the battery and the battery maximum available charge amount Q maxUpdate the matrix parameters in the battery extended Kalman filter model to calibrate the extended Kalman filter model, and estimate the state of charge of the battery according to the calibrated extended Kalman filter model.

[0133] Among them, the battery parameter set storage module includes:

[0134] The OCV-SOC relationship curve storage module is used to obtain the OCV-SOC data through intermittent charge and discharge experiments on the battery at a set of multiple temperatures, fit the data to obtain the expression of the OCV-SOC relationship curve at the corresponding temperature, store the expression of the OCV-SOC relationship curve obtained by fitting the data, and through this expression, the open circuit voltage value of the current battery can be calculated from the state of charge of the battery;

[0135] The expression of the OCV-SOC relationship curve is: OCV(t,SOC) = p1 t *SOC^7 + p2 t *SOC^6 + p3 t *SOC^5 + p4 t *SOC^4 + p5 t *SOC^3 + p6 t *SOC^2 + p7 t *SOC + p8 t

[0136] Among them, t is the set temperature, OCV is the open circuit voltage of the battery, SOC is the state of charge of the battery, the value range of SOC is 0 to 100%, p1 t ~p8 t are fitting coefficients;

[0137] The resistance-capacitance parameter relationship curve storage module is used to perform HPPC tests on the battery at different states of charge at the set temperature, obtain the corresponding battery response terminal voltage curve at this temperature and process it to obtain the corresponding resistance-capacitance parameters, and fit the resistance-capacitance parameters to obtain the corresponding resistance-capacitance parameter relationship curve and store it.

[0138] The battery maximum available charge amount update module is specifically:

[0139] According to Calculate, where Q passed is the charge amount flowing between the state of charge SOC 1 of the battery and the state of charge SOC 2 of the battery, SOC 1 and SOC 2 are the states of charge of the battery at two different moments when charging or discharging the battery, and SOC 1 and SOC 2The change in battery voltage needs to satisfy where dt is the time difference between two different moments, and dV is the voltage difference within the corresponding time difference.

[0140] The battery spatial state model establishment module includes:

[0141] The state equation establishment module: used to establish the state equation, and the state equation is:

[0142]

[0143] The observation equation establishment module, used to establish the observation equation, and the observation equation is:

[0144] U(K) = OCV(T, SOC) - U s (T, K) - U d (T, K) - R 0 (T, SOC)I(K) + ν K

[0145] where U s is the concentration polarization voltage of the battery, U d is the electrochemical polarization voltage of the battery, K is the state of charge of the battery calculated for the Kth time, τ s (T, SOC) = R s (T, SOC) * C s (T, SOC), τ d (T, SOC) = R d (T, SOC) * C d (T, SOC), I(K) is the battery current during the Kth cycle, Δt is the sampling time interval, η is the Coulomb efficiency, ω K is the process noise, ν K is the observation noise, e is a constant, R s is the concentration polarization resistance of the battery, R d is the electrochemical polarization resistance of the battery, C s is the concentration polarization capacitance of the battery, C d is the electrochemical polarization capacitance of the battery.

[0146] The extended Kalman filter model establishment module includes:

[0147] The state transition matrix establishment module, used to establish the state transition matrix, and the state transition matrix is:

[0148]

[0149] The input matrix establishment module, used to establish the input matrix, and the input matrix is:

[0150]

[0151] Observation matrix establishment module, used to establish an observation matrix, and the observation matrix is:

[0152]

[0153] Update module, used to update the parameters of the matrix in the extended Kalman filter model after the battery temperature T changes. The steps for the extended Kalman filter model to predict the state of charge of the battery and update the parameters in the matrix are as follows:

[0154] S4.1. Calculate the estimated state variable X K =(SOC(T, K + 1) U s (T, K + 1) U d (T, K + 1)) T ,

[0155] S4.2. Calculate the observed output value

[0156] S4.3. Update the parameters to solve the new state transition matrix A K-1 ,

[0157] S4.4. Calculate the pre-estimated value of the error covariance

[0158] S4.5. Update the parameters to find the new observation matrix C K ,

[0159] S4.6. Calculate the Kalman gain L K ,

[0160] S4.7. Calculate the optimal estimate of the state variable

[0161] S4.8. Calculate the optimal estimate of the error covariance

[0162] Among them, X 0 =(0.8 0.01 0.01) T , X 0 and are respectively the initial values of the algorithm iteration, R K is the observation noise error covariance, y Kis the battery terminal voltage measured at the K-th step, is the voltage prediction value of the current battery, Q K is the process noise.

Claims

1. A method for estimating the state of charge of a battery based on battery parameter compensation, characterized in that, it includes the following steps: S1. Based on the second-order RC equivalent circuit model of the battery to be evaluated, establish a data set of battery parameters to be evaluated. The data set of battery parameters to be evaluated includes the OCV-SOC relationship curve at different temperatures and the resistance-capacitance parameter relationship curve of the battery model obtained by HPPC tests at different temperatures and different states of charge; The method for obtaining the OCV-SOC relationship curve at different temperatures is as follows: After obtaining OCV-SOC data by performing intermittent charge and discharge experiments on the battery at a set of multiple temperatures, fit the data to obtain the expression of the OCV-SOC relationship curve at the corresponding temperature. Through this expression, the open-circuit voltage value of the current battery can be calculated using the state of charge of the battery. The expression of the OCV-SOC relationship curve is: OCV(t,SOC) = p1 t *SOC^7 + p2 t *SOC^6 + p3 t *SOC^5 + p4 t *SOC^4 + p5 t *SOC^3 + p6 t *SOC^2 + p7 t *SOC + p8 t where t is the set temperature, OCV is the open circuit voltage of the battery, SOC is the state of charge of the battery, the value range of SOC is 0 to 100%, and p1 t ~p8 t are fitting coefficients; The method for obtaining the resistance-capacitance parameter relationship curve of the battery model obtained by HPPC tests at different states of charge is as follows: Perform HPPC tests on the battery at different states of charge at a set temperature, obtain the corresponding battery response terminal voltage curve at this temperature and process it to obtain the corresponding resistance-capacitance parameters, and fit the resistance-capacitance parameters to obtain the corresponding resistance-capacitance parameter relationship curve; S2. Monitor the battery charge and discharge data to update the maximum available charge amount Q of the battery max ; The method of monitoring the charge and discharge data of the battery to update the maximum available charge Q of the battery max is as follows: According to calculate, where Q passed is the state of charge SOC of the battery 1 to the state of charge SOC of the battery 2 is the amount of charge flowing between them, SOC 1 and SOC 2 are the states of charge of the battery at two different times when charging or discharging the battery, and SOC 1 and SOC 2 require the battery voltage change to satisfy where dt is the time difference between two different times, and dV is the voltage difference within the corresponding time difference; S3. Based on the second-order RC equivalent circuit model of the battery, use the battery parameter data set and the maximum available charge Q max to establish a battery spatial state model; S4. Determine the battery extended Kalman filter model based on the battery space state model; update the matrix parameters in the battery extended Kalman filter model according to the real-time temperature of the battery and the maximum available charge Q of the battery max to correct the extended Kalman filter model, and estimate the state of charge of the battery according to the corrected extended Kalman filter model.

2. The method for estimating the state of charge of a battery based on battery parameter compensation according to claim 1, characterized in that, the battery state space model includes a state equation and an observation equation. The state equation is: The observation equation is: U(K) = OCV(T, SOC) - U s (T, K) - U d (T, K) - R 0 (T, SOC)I(K) + ν K Among them, U s is the concentration polarization voltage of the battery, U d is the electrochemical polarization voltage of the battery, K is the state of charge of the battery calculated for the Kth time, τ s (T, SOC) = R s (T, SOC) * C s (T, SOC), τ d (T, SOC) = R d (T, SOC) * C d (T, SOC), I(K) is the battery current during the Kth cycle, Δt is the sampling time interval, η is the Coulomb efficiency, ω K is the process noise, ν K is the observation noise, e is a constant, R s is the concentration polarization resistance of the battery, R d is the electrochemical polarization resistance of the battery, C s is the concentration polarization capacitance of the battery, C d is the electrochemical polarization capacitance of the battery, Q max is the maximum available charge of the battery; β 1 and β 2 are both representations of simplified parameters in the state equation; SOC(T, K) is the state of charge of the battery at the Kth cycle when the temperature is T; SOC(T, K + 1) is the state of charge of the battery at the (K + 1)th cycle when the temperature is T; R 0 (T, SOC) is the resistance of the battery when the temperature is T and the state of charge of the battery is SOC; τ s (T, SOC) is the concentration polarization time constant when the temperature is T and the state of charge of the battery is SOC; τ d (T, SOC) is the electrochemical polarization time constant when the temperature is T and the state of charge of the battery is SOC; R s (T, SOC) is the concentration polarization resistance when the temperature is T and the state of charge of the battery is SOC; OCV(T, SOC) is the open-circuit voltage of the battery when the temperature is T and the state of charge of the battery is SOC; C d (T, SOC) is the electrochemical polarization capacitance when the temperature is T and the state of charge of the battery is SOC.

3. The method for estimating the state of charge of a battery based on battery parameter compensation according to claim 2, characterized in that, the battery extended Kalman filter model includes a state transition matrix, an input matrix and an input matrix: The state transition matrix is: The input matrix is: The observation matrix is: After the battery temperature T changes, the parameters of the matrix in the extended Kalman filter model are updated. The steps for the extended Kalman filter model to predict the state of charge of the battery and update the parameters in the matrix are as follows: S4.

1. Calculate the estimated state variables X K = (SOC(T, K + 1)U s (T, K + 1)U d (T, K + 1)) T , is the optimal estimate of the state variables at the (K - 1)-th iteration; S4.

2. Calculate the observed output value is the first row element of the state variable prediction and is the second row element of the state variable prediction and is the third row element of the state variable prediction ; S4.

3. Update the parameters to solve the new state transition matrix A K-1 , For the estimation of state variables The first row elements S4.

4. Calculate the pre-estimated value of the error covariance is the optimal estimate of the error covariance for the (K-1)th iteration; S4.

5. Update the parameter to obtain a new observation matrix C K , S4.

6. Calculate the Kalman gain L K , S4.

7. Calculate the optimal estimate of the state variable S4.

8. Calculate the optimal estimate of the error covariance Among them, X 0 =(0.8 0.01 0.01) T , X 0 and are respectively the initial values of the algorithm iteration, R K is the observation noise error covariance, y K is the battery terminal voltage measured at the Kth step, is the voltage prediction value of the current battery, Q K is the process noise.

4. A system for estimating the state of charge of a battery based on battery parameter compensation, characterized in that, it includes: A battery parameter set storage module, which is used to establish a data set of battery parameters to be evaluated according to the second-order RC equivalent circuit model of the battery to be evaluated. The data set of battery parameters to be evaluated includes the OCV-SOC relationship curve at different temperatures and the resistance-capacitance parameter relationship curve of the battery model obtained by HPPC tests at different temperatures and different states of charge; The battery parameter set storage module includes: An OCV-SOC relationship curve storage module, which is used to obtain OCV-SOC data by performing intermittent charge and discharge experiments on the battery at a set of multiple temperatures, fit the data to obtain the expression of the OCV-SOC relationship curve at the corresponding temperature, and store the expression of the OCV-SOC relationship curve obtained by fitting the data. Through this expression, the open-circuit voltage value of the current battery can be calculated through the state of charge of the battery; The expression of the OCV-SOC relationship curve is: OCV(t,SOC) = p1 t *SOC^7 + p2 t *SOC^6 + p3 t *SOC^5 + p4 t *SOC^4 + p5 t *SOC^3 + p6 t *SOC^2 + p7 t *SOC + p8 t where t is the set temperature, OCV is the open-circuit voltage of the battery, SOC is the state of charge of the battery, the value range of SOC is 0 to 100%, and p1 t ~p8 t are fitting coefficients; The resistance-capacitance parameter relationship curve storage module is used to perform HPPC tests on the battery at different states of charge at a set temperature, obtain the corresponding battery response terminal voltage curve at this temperature and process it to obtain the corresponding resistance-capacitance parameters, fit the resistance-capacitance parameters to obtain the corresponding resistance-capacitance parameter relationship curve and store it; The maximum available charge amount update module of the battery is used to monitor the charge and discharge data of the battery to update the maximum available charge amount Q of the battery max ; The specific battery maximum available charge amount update module is as follows: According to calculate, where Q passed is the state of charge SOC of the battery 1 to the state of charge SOC of the battery 2 is the amount of charge flowing between them, SOC 1 and SOC 2 are the states of charge of the battery at two different times when charging or discharging the battery, and SOC 1 and SOC 2 require the battery voltage change to satisfy where dt is the time difference between two different times, and dV is the voltage difference within the corresponding time difference; The battery space state model establishment module is used to establish a battery space state model according to the second-order RC equivalent circuit model of the battery, using the battery parameter data set and the maximum available charge Q max ; The Kalman filter model establishment module is used to determine the battery extended Kalman filter model according to the battery space state model; and update the matrix parameters in the battery extended Kalman filter model based on the real-time temperature of the battery and the maximum available charge amount Q of the battery max to correct the extended Kalman filter model, and estimate the state of charge of the battery according to the corrected extended Kalman filter model.

5. The battery state of charge estimation system based on battery parameter compensation according to claim 4, characterized in that, The battery spatial state model establishment module includes: The state equation establishment module: used to establish the state equation, and the state equation is: The observation equation establishment module, used to establish the observation equation, and the observation equation is: U(K) = OCV(T, SOC) - U s (T, K) - U d (T, K) - R 0 (T, SOC)I(K) + ν K Among them, U s is the concentration polarization voltage of the battery, U d is the electrochemical polarization voltage of the battery, K is the state of charge of the battery calculated for the Kth time, τ s (T, SOC) = R s (T, SOC) * C s (T, SOC), τ d (T, SOC) = R d (T, SOC) * C d (T, SOC), I(K) is the battery current during the Kth cycle, Δt is the sampling time interval, η is the Coulomb efficiency, ω K is the process noise, ν K is the observation noise, e is a constant, R s is the concentration polarization resistance of the battery, R d is the electrochemical polarization resistance of the battery, C s is the concentration polarization capacitance of the battery, C d is the electrochemical polarization capacitance of the battery, Q max is the maximum available charge of the battery; β 1 and β 2 are both representations of simplified parameters in the state equation; SOC(T, K) is the state of charge of the battery at the Kth cycle when the temperature is T; SOC(T, K + 1) is the state of charge of the battery at the (K + 1)th cycle when the temperature is T; R 0 (T, SOC) is the resistance of the battery when the temperature is T and the state of charge of the battery is SOC; τ s (T, SOC) is the concentration polarization time constant when the temperature is T and the state of charge of the battery is SOC; τ d (T, SOC) is the electrochemical polarization time constant when the temperature is T and the state of charge of the battery is SOC; R s (T, SOC) is the concentration polarization resistance when the temperature is T and the state of charge of the battery is SOC; OCV(T, SOC) is the open circuit voltage of the battery when the temperature is T and the state of charge of the battery is SOC; C d (T, SOC) is the electrochemical polarization capacitance when the temperature is T and the state of charge of the battery is SOC.

6. The battery state of charge estimation system based on battery parameter compensation according to claim 5, characterized in that, The extended Kalman filter model establishment module includes: The state transition matrix establishment module, used to establish the state transition matrix, and the state transition matrix is: The input matrix establishment module, used to establish the input matrix, and the input matrix is: The observation matrix establishment module, used to establish the observation matrix, and the observation matrix is: The update module is used to update the parameters of the matrix in the extended Kalman filter model after the battery temperature T changes. The steps for the extended Kalman filter model to predict the battery state of charge and update the parameters in the matrix are as follows: S4.

1. Calculate the estimated state variables X K = (SOC(T, K + 1)U s (T, K + 1)U d (T, K + 1)) T , is the optimal estimate of the state variables in the (K - 1)-th iteration; S4.

2. Calculate the observed output value For the first row elements of the state variable prediction For the second row elements of the state variable prediction For the third row elements of the state variable prediction ​​​ S4.

3. Update the parameters to solve the new state transition matrix A K-1 , For the prediction of state variables The first row elements S4.

4. Calculate the pre-estimated value of the error covariance is the optimal estimate of the error covariance for the (K-1)-th iteration; S4.

5. Update the parameters to obtain a new observation matrix C K , S4.

6. Calculate the Kalman gain L K , S4.

7. Calculate the optimal estimate of the state variable S4.

8. Calculate the optimal estimate of the error covariance Among them, X 0 =(0.8 0.01 0.01) T , X 0 and are respectively the initial values of the algorithm iteration, R K is the observation noise error covariance, y K is the battery terminal voltage measured at the K-th step, is the voltage prediction value of the current battery, Q K is the process noise.

Images