A method for estimating the remaining capacity of a lithium-ion power battery

Abstract

The present invention discloses a method for estimating the remaining capacity of a lithium-ion power battery. First, the real-time sampled data and historical data are analyzed and processed, and the current SOC initial value is calculated according to the corresponding curve of the open circuit voltage and the SOC; then the SOC is estimated using the Kalman filter algorithm and the ampere-hour integration method algorithm respectively, and then the ampere-hour integration method is used as a reference to determine whether the Kalman filter result exceeds the tolerance of the SOC calculation error. Finally, if the error exceeds the tolerance, the SOC result is based on the ampere-hour integration method algorithm, and if the result does not exceed the error tolerance, the result is based on the Kalman filter calculation result. The present invention can ensure reliable and stable calculation of the battery remaining capacity under complex working conditions of electric vehicles, electric boats and hybrid ships, and improve the estimation accuracy of the battery remaining capacity over the entire life cycle of the system.

Description

Technical Field

[0001] The present invention belongs to the technical field of high-precision estimation of the remaining capacity of a lithium-ion power battery, and in particular relates to a method for estimating the remaining capacity of a lithium-ion power battery. Background Art

[0002] As one of the core algorithms of the battery management system, accurate SOC is a prerequisite for accurately characterizing the range of electric vehicles, electric ships, underwater vehicles, etc., and it is also a guarantee for the normal and effective operation of other functions of the battery management system.

[0003] Traditional methods for estimating the remaining power of power batteries mainly include the ampere-hour integration method and the open-circuit voltage method. However, the ampere-hour integration method is greatly affected by the accuracy of the current sensor, and the error accumulates over time. In order to solve the problem of low estimation accuracy of the state of charge of pure electric vehicle power batteries when the ambient temperature and discharge current change greatly, EKF is considered to be a more accurate algorithm for estimating the state of charge of electric vehicle power batteries. However, the essence of the EKF algorithm lies in the linear recursion of nonlinear functions. In order to ensure consistent convergence, the algorithm is mostly suitable for situations where the system nonlinearity is general and there is no strong coupling.

[0004] Given that the driving conditions of cars and ships are complex and unstable, as well as the nonlinear relationship between battery model parameters and battery capacity, a single remaining power estimation algorithm is not sufficient to match all operating conditions of the battery, and the overall estimation accuracy is low. Summary of the invention

[0005] The purpose of the present invention is to solve the problem that the driving state is complex and unstable, and the nonlinear relationship between the battery model parameters and the battery capacity leads to a single remaining power estimation algorithm that is insufficient to match all operating conditions of the battery, and the comprehensive estimation accuracy is low. An improved ampere-hour integration method and Kalman filtering are combined to estimate the remaining battery power.

[0006] The technical solution adopted by the present invention to solve the technical problem is: a method for estimating the remaining power of a lithium-ion power battery, comprising the following steps:

[0007] Step 1: After the battery management system is powered on for the first time, it starts sampling: the current battery temperature Tt of the battery is collected, and the current voltage Ut is used as the initial open circuit voltage OCV;

[0008] Step 2: Processing real-time sampling data and historical storage data: Analyze and count the collected current voltage Ut, current battery temperature Tt, current It, SOC value stored before startup, battery system static time and other information;

[0009] Step 3: Based on the initial open circuit voltage OCV and the current battery temperature Tt, combined with the historical SOC-OCV curve, the initial value SOC0 of the remaining battery capacity SOC corresponding to the initial open circuit voltage OCV is obtained for the first power-on of the algorithm, and the SOC-OCV curve is provided for the Kalman filter;

[0010] Step 4: Establish a battery equivalent circuit model and estimate the parameters of the battery equivalent circuit model based on the improved least squares method:

[0011] A second-order circuit model with two RC networks in series is selected. According to the electrical relationship of the battery equivalent circuit model, the current It is used as the input excitation, U is the output voltage, OCV is the open circuit voltage value, and y = OCV-U is the output response.

[0012]

[0013] In the formula, R0 is the ohmic internal resistance of the battery, R1 and C1 are the parameters of the first RC network, and R2 and C2 are the parameters of the second RC network. After solving, the following set of equations can be obtained:

[0014]

[0015] In the formula, the coefficients α1, α2, β0, β1, and β2 on the right side of the equation group have been obtained by the improved recursive least squares method. T is the sampling period, which is also a known quantity. Then, the coefficients R0, R1, C1, R2, and C2 on the left side can be obtained by solving the equation group.

[0016] Step 5: Use the extended Kalman filter algorithm to estimate the SOC value of the current system:

[0017] The Kalman filter equation is as follows

[0018] x k =Ax k-1 +Bu k-1 +w k-1

[0019] z k =Hx k +v k-1

[0020] In the formula, x k is the state variable of the system, z k is the observed variable of the system. The process excitation noise and observation noise of the system are represented by random signals w k With v k Indicates that u k represents the control function or system excitation; this formula is called the random state difference equation;

[0021] Combined with the second-order equivalent circuit model of the battery described in step 4, the various state parameters have the following relationship formula under the action of current excitation:

[0022]

[0023] In the formula, is the voltage across the ohmic internal resistance; is the voltage of the first RC network; is the voltage of the second RC network; C cap represents the capacity of the battery in ampere-seconds;

[0024] The state equation and observation equation are as follows:

[0025] x k =Ax k-1 +Bi k-1 +w k-1

[0026] z k =Hx k +v k-1

[0027]

[0028]

[0029]

[0030] H=[-1 -1 -1 h]

[0031]

[0032] In the formula, U EMF is the function of open circuit voltage with respect to SOC, h is the function of open circuit voltage with respect to U EMF About the derivative function of SOC;

[0033] Step 6, using the estimated SOC value calculated in the first 10 cycles of the extended Kalman filter algorithm as the initial value SOC0 of the ampere-hour integration method, and using the ampere-hour integration algorithm based on optimized compensation to estimate the current SOC value;

[0034] The calculation formula of the ampere-hour integral algorithm based on optimized compensation is as follows:

[0035]

[0036] In the formula, ω(τ) is the temperature correction coefficient, γ is the function of the number of battery cycles, η is the charge and discharge rate coefficient, and β is the battery aging and self-discharge correction coefficient. Substituting time t, i(t) into the formula, the estimated value of SOC at the current moment can be obtained:t ;

[0037] Step 7: Determine the SOC result of the Kalman filter algorithm k Whether the tolerance of SOC calculation error is exceeded △SOC: convert the SOC calculated by the ampere-hour integration method into t The estimated results are used as a benchmark to determine the SOC based on the extended Kalman filter algorithm. k Whether the estimated value exceeds the tolerance range of system error

[0038] |SOC k -SOC t |≤ΔSOC;

[0039] Step 8: If the result exceeds the SOC error tolerance ΔSOC, jump to step 9; if the result is within the error tolerance range, jump to step 10;

[0040] Step 9: Based on the result of the improved ampere-hour integration method, calculate the estimated SOC value of the current cycle SOC' = SOC t , jump to step 11;

[0041] Step 10: Based on the result of Kalman filter algorithm, calculate the estimated SOC value of the current cycle SOC' = SOC k , jump to step 11;

[0042] Step 11, using the calculated SOC estimation value SOC' as the current cycle SOC estimation value SOC=SOC', and updating the current cycle real-time sampling data;

[0043] Step 12: The system ends the task of estimating the remaining battery capacity based on the algorithm of combining ampere-hour integration and Kalman filtering.

[0044] Furthermore, the specific calculation process of SOC in step 5 is as follows:

[0045] The first step is to obtain the estimated value of the state variable at time k according to the state matrix equation:

[0046] x k * =Ax k-1 +Bi k-1 +w k-1

[0047] The asterisk indicates that the data is estimated based on the state equation. Accordingly, the covariance matrix after the state recursion is obtained:

[0048]

[0049] The second step is to solve the Kalman gain K k

[0050]

[0051] Among them, R k is the observation noise covariance matrix.

[0052] The third step is to modify the estimated value of the state vector and the corresponding covariance matrix according to the Kalman gain:

[0053] x k =x k * +K k (z k -Hx k * )=x k * +K k (u k -Hx k * )

[0054]

[0055] Here I is the unit matrix. In addition, since the working voltage u at both ends of the battery is used k As an observed variable, we have z k =u k ;

[0056] After the third step is completed, the time k increases by 1, and then the cycle returns to the first step to continue the calculation. At this point, the battery remaining capacity value estimation based on the battery model and the extended Kalman filter is realized, and the current SOC estimation value SOC k .

[0057] In the method for estimating the remaining capacity of a lithium-ion power battery, the SOC-OCV curve used for Kalman filtering in step 3 is a functional relationship between the open circuit voltage OCV and the SOC obtained when the battery is left to stand for a fixed time at different temperatures and different SOCs.

[0058] In the method for estimating the remaining capacity of a lithium-ion power battery, the Kalman filter algorithm in step five is used to estimate the remaining capacity of the battery. The Kalman filter of 10 cycles is used as the input value of the ampere-hour integration method, which can effectively optimize the accumulated error of the ampere-hour integration method and improve the global SOC estimation accuracy.

[0059] In the method for estimating the remaining capacity of a lithium-ion power battery, an improved ampere-hour integration algorithm is used in step six, and temperature, battery cycle times, charge and discharge rate, battery aging and self-discharge conditions are used as optimization parameters of the ampere-hour integration method.

[0060] The method for estimating the remaining capacity of a lithium-ion power battery adopts an ampere-hour integration algorithm. Aiming at the power consumption of electric vehicles, electric ships, hybrid electric vehicles, etc. under complex working conditions, the battery discharge rate will change suddenly, and the charging and discharging conditions will switch frequently, which will lead to the problem of Kalman filter divergence. By comparing the error between the ampere-hour integration and the Kalman filter algorithm within 10 seconds, if the difference is greater than the error tolerance, the ampere-hour integration is used to suppress the divergence of the Kalman filter, thereby improving the estimation accuracy of the remaining capacity of the local battery.

[0061] The beneficial effects of the present invention are:

[0062] The present invention adopts a combined estimation method of ampere-hour integration and Kalman filtering. Aiming at estimating the remaining capacity of the battery under complex working conditions, the combined estimation of the ampere-hour integration method and the Kalman filtering algorithm can suppress the divergence of the Kalman filtering while eliminating the cumulative error of the ampere-hour integration method. The estimation accuracy of the remaining capacity of the battery can reach more than 97%.

[0063] The present invention can ensure reliable and stable calculation of the remaining battery capacity under complex working conditions of electric vehicles, electric boats and hybrid ships, and improve the estimation accuracy of the remaining battery capacity over the entire life cycle of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] Figure 1 The figure is a flow chart of the method of the present invention. DETAILED DESCRIPTION

[0065] The present invention will be further described below in conjunction with the accompanying drawings and specific embodiments.

[0066] The present invention relates to a high-precision and high-stability battery remaining capacity estimation method. In order to solve the problem that the driving states of electric vehicles, electric ships and hybrid ships are complex and unstable, and the nonlinear relationship between battery model parameters and battery capacity leads to a single remaining capacity estimation algorithm being insufficient to match all battery working conditions, and the comprehensive estimation accuracy is low, an improved ampere-hour integration method and Kalman filtering are combined to estimate the remaining battery capacity.

[0067] The present invention aims at the equivalent model of lithium-ion power battery, adopts the improved ampere-hour integration method to optimize the SOC estimation algorithm of Kalman filter algorithm, and adopts the optimized filter algorithm to suppress the divergence of estimation error, so as to achieve the purpose of improving the estimation accuracy of battery remaining capacity.

[0068] First, the real-time sampled data and historical data are analyzed and processed, and the current SOC initial value is calculated based on the corresponding curve of open circuit voltage and SOC; then the SOC is estimated using the Kalman filter algorithm and the ampere-hour integration algorithm. Then, the ampere-hour integration method is used as a benchmark to determine whether the Kalman filter result exceeds the tolerance of the SOC calculation error; finally, if the error exceeds the tolerance, the SOC result is based on the ampere-hour integration algorithm, and if the result does not exceed the error tolerance, the result is based on the Kalman filter calculation result.

[0069] Specific implementation plan 1

[0070] Reference Figure 1 As shown, the present invention discloses a method for estimating the remaining power of a lithium-ion power battery, comprising the following steps:

[0071] Step 1: The battery management system starts sampling after being powered on for the first time: the current battery temperature Tt of the battery is collected, and the current voltage Ut is used as the initial open circuit voltage OCV.

[0072] Step 2: Processing real-time sampling data and historical storage data: Analyze and collect statistics on the collected current voltage Ut, current battery temperature Tt, current It, SOC value stored before startup, battery system static time and other information.

[0073] Step three, based on the initial open circuit voltage OCV and the current battery temperature Tt, combined with the historical SOC-OCV curve, the initial value SOC0 of the remaining battery capacity SOC corresponding to the initial open circuit voltage OCV is obtained for the first power-on of the algorithm, and the SOC-OCV curve is provided for the Kalman filter.

[0074] Step 4: Establish a battery equivalent circuit model and estimate the parameters of the battery equivalent circuit model based on the improved least squares method:

[0075] A second-order circuit model with two RC networks in series is selected. According to the electrical relationship of the battery equivalent circuit model, the current It is used as the input excitation, U is the output voltage, OCV is the open circuit voltage value, and y = OCV-U is the output response.

[0076]

[0077] In the formula, R0 is the ohmic internal resistance of the battery, R1 and C1 are the parameters of the first RC network, and R2 and C2 are the parameters of the second RC network. After solving, the following set of equations can be obtained:

[0078]

[0079] In the formula, the coefficients α1, α2, β0, β1, and β2 on the right side of the equation group have been obtained by the improved recursive least squares method. T is the sampling period, which is also a known quantity. The coefficients R0, R1, C1, R2, and C2 on the left side can be obtained by solving the equation group.

[0080] Step 5: Use the extended Kalman filter algorithm to estimate the SOC value of the current system:

[0081] The Kalman filter equation is as follows

[0082] x k =Ax k-1 +Bu k-1 +w k-1

[0083] z k =Hx k +v k-1

[0084] In the formula, x k is the state variable of the system, z k is the observed variable of the system. The process excitation noise and observation noise of the system are represented by random signals w k With v k Indicates that u k represents the control function or system excitation; this formula is called the random state difference equation;

[0085] Combined with the model of the second-order battery described in step 4, the various state parameters under the action of current excitation have the following relationship formula:

[0086]

[0087] In the formula, is the voltage across the ohmic internal resistance; is the voltage of the first RC network; is the voltage of the second RC network; C cap represents the capacity of the battery in ampere-seconds;

[0088] The state equation and observation equation are as follows:

[0089] x k =Ax k-1 +Bi k-1 +w k-1

[0090] z k =Hx k +v k-1

[0091]

[0092]

[0093]

[0094] H=[-1-1-1h]

[0095]

[0096] In the formula, U EMF is the function of open circuit voltage with respect to SOC, h is the function of open circuit voltage with respect to U EMF About the derivative function of SOC.

[0097] The specific calculation process of SOC is as follows:

[0098] The first step is to obtain the estimated value of the state variable at time k according to the state matrix equation:

[0099] x k * =Ax k-1 +Bi k-1 +w k-1

[0100] Here the asterisk indicates that the data is estimated based on the state equation. Accordingly, the covariance matrix after the state recursion is obtained:

[0101]

[0102] The second step is to solve the Kalman gain K k

[0103]

[0104] Among them, R k is the observation noise covariance matrix.

[0105] The third step is to modify the estimated value of the state vector and the corresponding covariance matrix according to the Kalman gain:

[0106] x k =x k * +K k (z k -Hx k * )=x k * +K k (u k -Hx k * )

[0107]

[0108] Here I is the unit matrix. In addition, since the working voltage u at both ends of the battery is used k As an observed variable, we have z k =u k .

[0109] After the third step is completed, the time k increases by 1, and then the cycle returns to the first step to continue the calculation. At this point, the battery remaining capacity value estimation based on the battery model and the extended Kalman filter is realized, and the current SOC estimation value SOC k .

[0110] Step 6, using the estimated SOC value calculated in the first 10 cycles of the extended Kalman filter algorithm as the initial value SOC0 of the ampere-hour integration method, and using the ampere-hour integration algorithm based on optimized compensation to estimate the current SOC value;

[0111] The calculation formula of the ampere-hour integral algorithm based on optimized compensation is as follows:

[0112]

[0113] In the formula, ω(τ) is the temperature correction coefficient, γ is the function of the number of battery cycles, η is the charge and discharge rate coefficient, and β is the battery aging and self-discharge correction coefficient. Substituting time t, i(t) into the formula, the estimated value of SOC at the current moment can be obtained: t .

[0114] Step 7: Determine the SOC result of the Kalman filter algorithm k Whether the tolerance of SOC calculation error is exceeded △SOC: convert the SOC calculated by the ampere-hour integration method into t The estimated results are used as a benchmark to determine the SOC based on the extended Kalman filter algorithm. k Whether the estimated value exceeds the tolerance range of system error

[0115] |SOC k -SOC t |≤ΔSOC.

[0116] Step 8: If the result exceeds the SOC error tolerance ΔSOC, jump to step 9; if the result is within the error tolerance range, jump to step 10.

[0117] Step 9: Based on the result of the improved ampere-hour integration method, calculate the estimated SOC value of the current cycle SOC' = SOC t , jump to step 11.

[0118] Step 10: Based on the result of Kalman filter algorithm, calculate the estimated SOC value of the current cycle SOC' = SOC k, jump to step 11.

[0119] Step eleven: taking the calculated SOC estimation value SOC' as the current period SOC estimation value SOC=SOC', and updating the current period real-time sampling data.

[0120] Step 12: The system ends the task of estimating the remaining battery capacity based on the algorithm of combining ampere-hour integration and Kalman filtering.

[0121] Specific implementation plan 2

[0122] This implementation method further explains step three of the algorithm for jointly estimating the remaining battery capacity by combining ampere-hour integration and Kalman filtering in the specific implementation method one, and is characterized by the SOC-OCV function, which is the functional relationship between the open circuit voltage OCV and SOC obtained when the battery is left standing for a fixed time at different temperatures and different SOCs.

[0123] Specific implementation plan three

[0124] This implementation further illustrates step five of the algorithm for estimating the remaining battery capacity by combining the ampere-hour integration and the Kalman filter in the specific implementation mode one, and is characterized in that the Kalman filter algorithm is used to estimate the remaining battery capacity, which can effectively optimize the accumulated error of the ampere-hour integration method and improve the global SOC estimation accuracy.

[0125] The Kalman filter algorithm is used to estimate the remaining capacity of the battery. The Kalman filter of 10 cycles is used as the input value of the ampere-hour integration method, which can effectively optimize the accumulated error of the ampere-hour integration method and improve the global SOC estimation accuracy.

[0126] Specific implementation plan four

[0127] This implementation manner further illustrates step six of the algorithm for jointly estimating the remaining battery capacity by combining the ampere-hour integration and the Kalman filter in the specific implementation manner one, and is characterized in that an improved ampere-hour integration algorithm is adopted, and temperature, battery cycle times, charge and discharge rates, battery aging and self-discharge conditions are used as optimization parameters of the ampere-hour integration method, which can effectively improve the estimation accuracy of the remaining battery capacity.

[0128] Specific implementation plan five

[0129] This embodiment further illustrates step 7 of the algorithm for estimating the remaining battery capacity by combining ampere-hour integration and Kalman filtering in the first embodiment, which is characterized by adopting the ampere-hour integration algorithm. For the power consumption of electric vehicles, electric ships, hybrid electric vehicles, etc. under complex working conditions, the battery discharge rate will suddenly change, and the charging and discharging conditions will also switch frequently, which will cause the Kalman filter to diverge. The participation of the ampere-hour integration method in the correction can effectively suppress the divergence of the Kalman filter and improve the accuracy of the estimation of the remaining capacity of the local battery.

[0130] Specific implementation plan six

[0131] The present embodiment further illustrates the algorithm for jointly estimating the remaining battery capacity by combining the ampere-hour integration and the Kalman filter of the specific embodiment one, and is characterized by adopting the combined estimation method of the ampere-hour integration and the Kalman filter. For estimating the remaining battery capacity under complex working conditions, the combined estimation of the ampere-hour integration method and the Kalman filter algorithm can suppress the divergence of the Kalman filter while eliminating the accumulated error of the ampere-hour integration method, and the estimation accuracy of the remaining battery capacity can reach less than 3%.

[0132] The ampere-hour integration algorithm is used to estimate the remaining battery capacity under complex working conditions for electric vehicles, electric ships, hybrid vehicles, etc. The battery discharge rate will change suddenly, and the charging and discharging conditions will switch frequently, which will lead to the problem of Kalman filter divergence. By comparing the errors between the ampere-hour integration and the Kalman filter algorithm within 10 seconds, if the difference is greater than the error tolerance, the ampere-hour integration is used to suppress the divergence of the Kalman filter and improve the estimation accuracy of the local battery remaining capacity. The ampere-hour integration and Kalman filter joint estimation method is used to estimate the remaining battery capacity under complex working conditions. The ampere-hour integration method and the Kalman filter algorithm can suppress the divergence of the Kalman filter while eliminating the accumulated error of the ampere-hour integration method. The estimation accuracy of the remaining battery capacity can reach more than 97%.

[0133] The present invention is not limited to the above-mentioned optimal implementation mode. Any technical personnel in this field can derive other modified and improved products under the enlightenment of the present invention. However, no matter what changes are made in its shape or structure, all technical solutions that are the same or similar to those of the present application fall within the protection scope of the present invention.

Claims

1. A method for estimating the remaining power of a lithium-ion power battery, characterized in that: The following steps are included Step 1: collect the current battery temperature Tt of the battery and use the current voltage Ut as the initial open circuit voltage OCV; Step 2, sampling the SOC value stored before startup; Step 3: Based on the initial open circuit voltage OCV and the current battery temperature Tt, combined with the historical SOC-OCV curve, the initial value SOC0 of the battery remaining capacity SOC corresponding to the initial open circuit voltage OCV is obtained, and the SOC-OCV curve for Kalman filtering is obtained; Step 4: Establish a battery equivalent circuit model and estimate the parameters of the battery equivalent circuit model based on the improved least squares method: A second-order circuit model with two RC networks in series is selected. According to the electrical relationship of the battery equivalent circuit model, the current It is used as the input excitation, U is the output voltage, OCV is the open circuit voltage value, and y = OCV-U is the output response. In the formula, R0 is the ohmic internal resistance of the battery, R1 and C1 are the parameters of the first RC network, and R2 and C2 are the parameters of the second RC network. After solving, the following set of equations can be obtained: In the formula, the coefficients α1, α2, β0, β1, and β2 on the right have been obtained by the improved recursive least squares method, T is the sampling period, and the coefficients R0, R1, C1, R2, and C2 on the left can be obtained by solving the equations; Step 5: Use the extended Kalman filter algorithm to estimate the SOC value of the current system: x k =Ax k-1 +Bu k-1 +w k-1 z k =Hx k +v k-1 In the formula, x k is the state variable of the system, z k is the observed variable of the system, w k With v k represents the process excitation noise and observation noise of the system, u k represents the control function or called system excitation; Combined with the battery second-order circuit model, the various state parameters have the following relationship formula under the action of current excitation: In the formula, is the voltage across the ohmic internal resistance; is the voltage of the first RC network; is the voltage of the second RC network; C cap represents the capacity of the battery; The state equation and observation equation are as follows: x k =Ax k-1 +Bi k-1 +w k-1 z k =Hx k +v k-1 H=[-1 -1 -1 h] In the formula, U EMF is the function of open circuit voltage with respect to SOC, h is the function of open circuit voltage with respect to U EMF About the derivative function of SOC; Step 6, using the estimated SOC value calculated in the first 10 cycles of the extended Kalman filter algorithm as the initial value SOC0 of the ampere-hour integration method, and using the ampere-hour integration algorithm based on optimized compensation to estimate the current SOC value; The calculation formula of the ampere-hour integral algorithm based on optimized compensation is as follows: In the formula, ω(τ) is the temperature correction coefficient, γ is the function of the number of battery cycles, η is the charge and discharge rate coefficient, and β is the battery aging and self-discharge correction coefficient. Substitute time t, i(t) into the formula to obtain the estimated value of SOC at the current moment: SOC t ; Step 7: Determine the SOC result of the Kalman filter algorithm k Whether the tolerance of SOC calculation error is exceeded △SOC: convert the SOC calculated by the ampere-hour integration method into t The estimated results are used as a benchmark to determine the SOC based on the extended Kalman filter algorithm. k Whether the estimated value exceeds the tolerance range of system error |SOC k -SOC t |≤ΔSOC; Step 8: If the result exceeds the SOC error tolerance △SOC, jump to step 9; If the result is within the error tolerance range, jump to step 10; Step 9: Calculate the current cycle SOC estimate SOC' = SOC using the improved ampere-hour integration method result t , jump to step 11; Step 10: Calculate the current cycle SOC estimation value SOC' = SOC using the result of the Kalman filter algorithm k , jump to step 11; Step 11, using the calculated SOC estimation value SOC' as the current cycle SOC estimation value SOC=SOC', and updating the current cycle real-time sampling data; Step 12: The system ends the task of estimating the remaining battery power.

2. A method for estimating the remaining power of a lithium-ion power battery according to claim 1, characterized in that: The specific calculation process of SOC in step 5 is as follows: The first step is to obtain the estimated value of the state variable at time k according to the state matrix equation: x k * =Ax k-1 +Bi k-1 +w k-1 The asterisk indicates that the data is estimated based on the state equation, and the covariance matrix after state recursion is obtained. The second step is to solve the Kalman gain In the formula, R k is the observation noise covariance matrix; The third step is to modify the estimated value of the state vector and the corresponding covariance matrix according to the Kalman gain: x k =x k * +K k (z k -Hx k * )=x k * +K k (she is k -Hx k * ) Where I is the unit matrix, and the working voltage u at both ends of the battery is used. k As the observed variable, z k =u k ; Loop the first step and continue calculating to get the current SOC estimate SOC k .

3. A method for estimating the remaining power of a lithium-ion power battery according to claim 1 or 2, characterized in that: The SOC-OCV curve used for Kalman filtering in step three is a functional relationship between the open circuit voltage OCV and the SOC obtained when the battery is left to stand for a fixed time at different temperatures and different SOCs.

4. A method for estimating the remaining power of a lithium-ion power battery according to claim 1 or 2, characterized in that: When the Kalman filter algorithm in step 5 estimates the remaining capacity of the battery, a 10-cycle Kalman filter is used as an input value of the ampere-hour integration method.

5. A method for estimating the remaining power of a lithium-ion power battery according to claim 1 or 2, characterized in that: In step six, an improved ampere-hour integration algorithm is used, and temperature, battery cycle times, charge and discharge rate, battery aging and self-discharge conditions are used as optimization parameters of the ampere-hour integration method.

6. A method for estimating the remaining power of a lithium-ion power battery according to claim 1, characterized in that: When the ampere-hour integration algorithm is used, the error between the ampere-hour integration and the Kalman filter algorithm within 10 seconds is compared. If the difference is greater than the error tolerance, the ampere-hour integration is used to suppress the divergence of the Kalman filter.

Images