Lithium iron phosphate battery SOC estimation method considering low temperature influence

Abstract

The invention discloses a lithium iron phosphate battery SOC (State of Charge) estimation method considering low-temperature influence, which comprises the following steps of: firstly, acquiring voltage and current data of a temperature gradient in a low-temperature range through a low-current constant-current charge-discharge experiment, and calculating an SOC value and a discharge capacity C of each sampling point; secondly, according to data of each sampling point, adopting interpolation, spline interpolation and polynomial fitting methods to obtain all OCV-SOC curves and a curve of discharge capacity changing along with temperature in a low-temperature range; and finally, performing parameter identification by using a self-adaptive recursive least square method and calculating a real-time SOC value by using a self-adaptive extended Kalman filtering algorithm. According to the invention, based on the influence of low temperature on the lithium iron phosphate battery, the SOC estimation method is improved by correcting the OCV-SOC curve and the discharge capacity C, so that the estimation precision of the SOC is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of lithium battery management, and particularly relates to a method for estimating the state of charge (SOC) of a lithium iron phosphate battery considering the influence of low temperature. Background Art

[0002] In applications such as electric vehicles and energy storage systems, the estimation of the state of charge (SOC) of lithium-ion batteries is a key factor to ensure the safe and efficient operation of the system. Lithium iron phosphate (LiFePO4) batteries have been widely used due to their excellent safety, good thermal stability, and long cycle life, especially in the fields of electric vehicles and renewable energy storage. However, the influence of low temperature on battery performance cannot be ignored, especially in terms of the accuracy of SOC estimation.

[0003] Under low temperature conditions, the electrochemical reaction rate of lithium iron phosphate batteries decreases significantly, and the internal resistance increases, which leads to a decrease in available capacity and a weakening of energy output. This situation makes the estimation of SOC vulnerable to influence and generates large errors. This is because these methods usually assume that the battery operates at standard temperature and do not fully consider the influence of low temperature on battery chemical reactions and performance. Therefore, low temperature correction is particularly important for the SOC estimation of lithium iron phosphate batteries. It can not only improve the adaptability of the battery in harsh environments, but also provide strong support for the optimization of the battery management system and the safety performance of electric vehicles. Therefore, in-depth research and development of low temperature correction technology will provide an important guarantee for the reliability and efficiency of lithium iron phosphate batteries in practical applications. To improve the accuracy of SOC estimation in low temperature environments, many scholars have proposed various low temperature correction technologies. However, these correction methods usually only target a few specific temperatures and cannot cover all low temperature ranges. Moreover, the influence of low temperature on the charge and discharge performance of lithium iron phosphate batteries is non-linear, and a small temperature change will cause a large change in the open circuit voltage. It is obviously unrealistic to obtain data through experiments for all low temperature ranges. Summary of the Invention

[0004] The purpose of the present invention is to provide a method for estimating the state of charge (SOC) of a lithium iron phosphate battery considering the influence of low temperature.

[0005] The technical solution for achieving the purpose of the present invention is as follows: A method for estimating the state of charge (SOC) of a lithium iron phosphate battery considering the influence of low temperature, comprising the following steps:

[0006] Step 1, through low-current constant-current charge and discharge experiments, obtain voltage and current data in low temperature range gradients, and calculate the remaining battery charge value SOC and discharge capacity C at each sampling point;

[0007] Step 2: Synchronize the SOC sampling intervals at each temperature gradient using the interpolation method. Obtain new data points by taking the mean of the charge-discharge curves, and use polynomial fitting to obtain the OCV-SOC curve in the low-temperature range. Use the spline interpolation method to obtain the OCV values with a temperature interval of 1°C at different SOCs, and use polynomial fitting to obtain the OCV-SOC curves in each low-temperature range. Fit the C-T curve by polynomial to obtain the discharge capacity C at each temperature.

[0008] Step 3: Establish the battery equivalent circuit model and state equation, use the adaptive recursive least squares method for online parameter identification, and then calculate the real-time SOC value based on the adaptive extended Kalman filter algorithm.

[0009] 2. The SOC estimation method for low-temperature correction of lithium iron phosphate batteries according to claim 1, wherein in step 1, through low-current constant current charge-discharge experiments, obtain the voltage and current data of each temperature gradient in the low-temperature range, and calculate the remaining battery charge value SOC and discharge capacity C of each sampling point. The specific method is as follows:

[0010] 1) Discharge at a constant current C until the battery is fully discharged.

[0011] 2) Adjust the temperature of the thermostat, put the lithium battery into the thermostat and let it stand for 30 minutes.

[0012] 3) Charge the battery according to the standard charging protocol until the current drops to 0 in the constant voltage mode, and collect the voltage U and current I during the charging process.

[0013] 4) Discharge at a constant rate of C / 20 until the battery is fully discharged, and collect the voltage and current data during the discharge process.

[0014] 5) In the range of 10°C to -40°C, collect the voltage (U) and current (I) data of each temperature gradient in the low-temperature range at intervals of 5°C.

[0015] 6) Calculate the remaining battery charge SOC value and discharge capacity C of each sampling point.

[0016] Furthermore, in step 2, synchronize the SOC sampling intervals at each temperature gradient using the interpolation method. Obtain new data points by taking the mean of the charge-discharge curves, and use polynomial fitting to obtain the OCV-SOC curve in the low-temperature range. Use the spline interpolation method to obtain the OCV values with a temperature interval of 1°C at different SOCs, and use polynomial fitting to obtain the OCV-SOC curves in each low-temperature range. Fit the C-T curve by polynomial to obtain the discharge capacity C at each temperature. The specific method is as follows:

[0017] 2.1) Synchronize the SOC sampling intervals at each temperature gradient using the interpolation method. Obtain new data points by taking the average of the charge and discharge curves, and use polynomial fitting to obtain the OCV-SOC curve in the low-temperature range;

[0018] 2.1.1) Based on the state of charge (SOC) of the battery and the voltage (U) in the charge and discharge experiments, use the interpolation method to uniformly divide the charge and discharge data points (SOC, U) at all gradient temperatures into 1000 each, and calculate the values of the interpolation points: First, calculate the slope Then calculate the OCV of the point where interpolation is required C = OCV0 + k(SOC - SOC0), where OCV1 is the voltage corresponding to the subsequent SOC point, OCV0 is the voltage corresponding to the previous SOC point, SOC1 and SOC0 are the two adjacent points of the point where interpolation is required, and OCV C is the interpolated voltage obtained in the middle, and SOC is the point corresponding to the middle interpolation;

[0019] 2.1.2) Considering the hysteresis characteristic of the open-circuit voltage, take the average of the charging interpolation voltage and the discharging interpolation voltage OCV = (OCV C + OCV D ) / 2, where OCV C represents the charging interpolation voltage, and OCV D represents the discharging interpolation voltage;

[0020] 2.1.3) Use polynomial fitting to obtain the OCV-SOC curve corresponding to the gradient temperature OCV = a0 + a1SOC + a2SOC 2 + a3SOC 3 + a4xSOC 4 + a5SOC 5 + a6SOC 6 + a7SOC 7 +

[0021] a8SOC 8 + a9SOC 9 + a 10 SOC 10 + a 11 SOC 11 + a 12 SOC 12 , where a0, a1, …, a 12 are coefficients to be determined;

[0022] 2.1.4) From 10 °C to -40 °C, at intervals of 5 °C, loop through steps 2.1 to 2.3 until all OCV-SOC curves in the low-temperature range are fitted;

[0023] 2.2) Use the spline interpolation method to obtain the OCV values at a temperature interval of 1 °C for different SOCs, and use polynomial fitting to obtain the OCV-SOC curves in each low-temperature range. The specific method is as follows:

[0024] 2.2.1) Take out the OCV values at different temperatures corresponding to the same SOC value at intervals of 5 °C. The value range of SOC is from 0 to 1, and the sampling interval is 0.05;

[0025] 2.2.2) Use cubic spline interpolation to obtain all the corresponding OCV values in the low-temperature range. The steps of spline interpolation are as follows: First, for the given data point sequence (T0, OCV0), (T1, OCV1), (T2, OCV2),…, (T 10 , OCV 10 ), T X is the measured temperature, and OCV X is the open-circuit voltage corresponding to the SOC and temperature at this time. Calculate the distance h i = T i+1 - T i (i = 0, 1,…, 9). The parameters to be calculated are as follows:

[0026]

[0027] where μ i , λ i , d i are the relevant coefficients for constructing the equations.

[0028]

[0029] Solve for M i . For each pair of adjacent data points (T i , OCV i ) and (T i+1 , OCV i+1 ) (i = 0, 1,…, 9), perform the following operations:

[0030]

[0031] where T i1 , T i2 , T i3 , T i4 are the temperatures of the 4 inserted points, and they are evenly distributed in the interval [T i , T i+1 . Calculate all the OCV values at each temperature interval for the same SOC value:

[0032]

[0033] where

[0034] After calculating all OCV values at the same SOC value from -10°C to 40°C with an interval of 1°C, the SOC value is rolled and changed, and then the SOC, T, OCV combinations of all groups with an interval of 0.05 from 0 to 1 for SOC are obtained;

[0035] 2.2.3) Take out the data points at the same temperature T, with the abscissa being SOC and the ordinate being OCV, and use a 12th-order polynomial fitting to obtain the OCV-SOC curve at intervals of 1°C within each low-temperature range;

[0036] 2.3) Fit the C-T curve through polynomial fitting to obtain the discharge capacity C at each temperature. The specific method is as follows:

[0037] Take the temperature T as the abscissa and the discharge capacity C as the ordinate, fit the C-T curve using a 3rd-order polynomial, and use the 3rd-order polynomial function to estimate the discharge capacity C corresponding to all temperatures at intervals of 1°C within the low-temperature range.

[0038] Furthermore, in step 3, establish a battery equivalent circuit model and state equation, use the adaptive recursive least squares method for online parameter identification, and then calculate the real-time SOC value based on the adaptive extended Kalman filter algorithm. The specific method is as follows:

[0039] 3.1) Establish a second-order RC equivalent circuit model according to the battery, including a voltage source representing the open-circuit voltage U ocv , a series resistance R0, and two parallel RC networks, namely R1, C1 and R2, C2, which are expressed as:

[0040]

[0041] where U ocv is the open-circuit voltage, obtained from the SOC and OCV-SOC curve at the previous moment. R0 is the ohmic internal resistance, R1 and C1 are the electrochemical polarization resistance and capacitance respectively, R2 and C2 are the concentration polarization resistance and capacitance respectively, U1 and U2 are the electrochemical polarization voltage and concentration polarization voltage respectively, U is the voltage across the circuit, that is, the measured voltage of the actual operation of the battery, and I is the current across the circuit, that is, the measured current of the actual operation of the battery;

[0042] 3.2) Use the adaptive recursive least squares method for parameter identification to identify R0, R1, C1, R2, C2 in the equivalent circuit. Specifically:

[0043] 3.2.1) Initialize, set the initial parameter estimation value and the initial covariance matrix P0;

[0044] 3.2.2) Calculate the predicted output y(k) based on the current parameter estimates and the input :

[0045]

[0046] where

[0047] 3.2.3) Calculate the gain matrix K k , update the parameter estimates and update the covariance matrix P k :

[0048]

[0049] where y(k) = U ocv (k) - U(k);

[0050] 3.2.4) Update the dynamic genetic factor λ k :

[0051] λ k = α + (1 - α)e -γεk (10)

[0052] where α is an adjustable parameter less than 1 and close to 1; γ is a positive adaptation coefficient;

[0053] 3.2.5) Perform parameter identification and convert to R0, R1, R2, C1, C2:

[0054]

[0055]

[0056] where T is the sampling period, a = τ1τ2, b = τ1 + τ2, loop from 3.2.2) to 3.2.5) until all data is identified; then take the mean values of R0, R1, R2, C1, C2 respectively to obtain the input-adaptive extended Kalman filter algorithm module;

[0057] 3.3) Calculate the real-time SOC value using the adaptive extended Kalman filter algorithm, specifically:

[0058] 3.3.1) Establish the state equation and observation equation based on the battery equivalent circuit model:

[0059]

[0060] U k = -U 1,k -U2,k +U OCV,k -I k R0 (14)

[0061] where η is the charge-discharge efficiency, Q is the actual discharge capacity of the battery, and U 1,k is the voltage across R1 at time K, and U 2,k is the voltage across R2 at time K. τ1 and τ2 have the same equivalent relationship as in Equation (12), and Ts is the sampling period;

[0062] 3.3.2) Initialize the state vector Initialize the covariance matrix P0, set the process noise covariance matrix Q0 and the measurement noise covariance R0, and initialize the Jacobian matrix H0 = [-1 -1 0]. These matrices are initially estimated and set according to the noise characteristics of the system;

[0063] 3.3.3) Make a prediction and perform a prior state prediction for the next moment according to the state equation Calculate the open-circuit voltage U using the prior estimate and the OCV-SOC curve corresponding to the temperature OCV,k , and update the observation matrix H through k = [-1 -1 U OCV,k , and estimate the terminal voltage using the observation equation Perform a prior state covariance prediction:

[0064]

[0065]

[0066] 3.3.4) Update step, calculate the Kalman gain:

[0067]

[0068] Update the posterior state value:

[0069]

[0070] where U k is the measured terminal voltage value, and update the error covariance matrix:

[0071] P k = (I е - K k H k )P k / k-1 (19)

[0072] where I е is the identity matrix, adaptively adjust the measurement noise covariance R k and the process noise covariance matrix Q k :

[0073]

[0074] wherein α and β represent adjustment factors, and e k is the error between the measured terminal voltage and the estimated terminal voltage.

[0075] Loop through steps 3.3.3) to 3.3.4) until all data is predicted, and output the state vectors at all times Extract all SOC values.

[0076] An SOC estimation system for lithium iron phosphate batteries considering low-temperature effects, implementing the above-mentioned SOC estimation method for lithium iron phosphate batteries considering low-temperature effects, to achieve SOC estimation of lithium iron phosphate batteries considering low-temperature effects, and separately execute steps 1 - 3 in three modules.

[0077] A computer device includes a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, it implements the above-mentioned SOC estimation method for lithium iron phosphate batteries considering low-temperature effects, to achieve SOC estimation of lithium iron phosphate batteries considering low-temperature effects.

[0078] A computer-readable storage medium stores a computer program. When the computer program is executed by a processor, it implements the above-mentioned SOC estimation method for lithium iron phosphate batteries considering low-temperature effects, to achieve SOC estimation of lithium iron phosphate batteries considering low-temperature effects.

[0079] Compared with the prior art, the significant advantages of the present invention are as follows: By using a relatively small experimental cost, covering all low-temperature ranges, establishing a relationship model between battery performance and temperature, so as to dynamically adjust the SOC estimation. By correcting the relationship between the battery voltage and SOC, the accuracy and reliability of the SOC estimation are enhanced, enabling the battery management system to better adapt to different temperature conditions and providing more accurate state feedback. Description of the Drawings

[0080] Figure 1 is a schematic flow diagram of SOC estimation.

[0081] Figure 2 is a schematic diagram of solving the OCV - SOC curve.

[0082] Figure 3 is a schematic diagram of solving interpolation.

[0083] Figure 4 is a schematic diagram of fitting the OCV - SOC curve.

[0084] Figure 5It is a second-order equivalent circuit model.

[0085] Figure 6 It is a comparison example diagram of temperature correction and non-correction. Detailed implementation manners

[0086] In order to make the objectives, technical solutions and advantages of the present application clearer and more understandable, the present application will be further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0087] A method for estimating the SOC of a lithium iron phosphate battery considering the influence of low temperature, the general idea is: obtain the OCV-SOC curve and C-T curve through a low-current constant-current experiment, and input the terminal voltage U, current I, open-circuit voltage OCV (obtained from the SOC and OCV-SOC curve at the previous moment), and discharge capacity C (obtained from the C-T curve, and the temperature T is a known quantity) during the actual operation of the lithium battery (other working conditions) into the AFFRLS (adaptive recursive least squares) algorithm module for battery model parameter identification to obtain the parameters R0, R1, R2, C1, C2, and then input these parameters and the actual operation data (terminal voltage U, current I, open-circuit voltage OCV, discharge capacity C) into the AEKF (adaptive extended Kalman filter) module to obtain the SOC of the lithium battery at this time (and synchronously use the OCV-SOC curve to calculate the OCV at this time for parameter identification at the next moment, and so on), so as to realize the estimation of the SOC of the lithium battery. The specific operations are as follows:

[0088] The first step is to obtain the OCV-SOC curve and discharge capacity in a ladder (interval of 5 °C) in the low-temperature range (-40 °C to 10 °C) through a low-current constant-current charge and discharge experiment. The specific steps are as follows:

[0089] 1) Discharge at a constant current of C until the battery is fully discharged; 2) Adjust the temperature of the thermostat (from 10 °C to -40 °C, at intervals of 5 °C), put the lithium battery into the thermostat and let it stand for 30 minutes; 3) Charge the battery according to the standard charging protocol until the current drops to 0 in the constant voltage mode, and collect the voltage and current data during the charging process;

[0090] 4) Discharge at a constant rate of C / 20 until the battery is fully discharged, and collect the voltage and current data during the discharge process;

[0091] 5) Repeat the experiment according to steps 2) to 4) until all data from 10 °C to -40 °C at intervals of 5 °C are collected;

[0092] 6) Calculate the SOC value and discharge capacity C of each sampling point.

[0093] Step 2: First, use the interpolation method to synchronize the SOC sampling intervals at each temperature gradient (i.e., align the data points, with the abscissa of the data points being SOC and the ordinate being the voltage U in the charge-discharge experiment). Obtain new data points by taking the average of the charge-discharge curves (the abscissa of the data points is SOC and the ordinate is the open-circuit voltage OCV). Use polynomial fitting to obtain the OCV-SOC curve in the low-temperature range (interval of 5 °C). The specific method is as follows:

[0094] 2.1) Use the interpolation method to evenly divide the charge data points at all gradient temperatures into 1000 each, and calculate the values of the interpolation points: First, calculate the slope Then calculate the OCV of the point where interpolation is required C = OCV0 + k(SOC - SOC0), where OCV1 is the voltage corresponding to the subsequent SOC point, OCV0 is the voltage corresponding to the previous SOC point, SOC1 and SOC0 are the two adjacent points of the point where interpolation is required, and OCV C is the interpolated voltage obtained in the middle, and SOC is the point corresponding to the middle interpolation; Use the interpolation method to evenly divide the discharge data points at all gradient temperatures into 1000 each, and calculate the values of the interpolation points. The method is the same as the charge data processing:

[0095] 2.2) Considering the hysteresis characteristic of the open-circuit voltage, take the average value of the 1000 data points of charge and discharge respectively (the abscissa is SOC and the ordinate is OCV, which can be higher according to the accuracy requirements), that is, the charge interpolation voltage and the discharge interpolation voltage obtained in step 2.1, OCV = (OCV C + OCV D ) / 2, where OCV C represents the charge interpolation voltage, and OCV D represents the discharge interpolation voltage;

[0096] 2.3) Use polynomial fitting for these 1000 data points (the ordinate is the average value OCV obtained in step 2.2, and the abscissa is the SOC obtained by corresponding interpolation), OCV = a0 + a1SOC + a2SOC 2 + a3SOC 3 + a4xSOC 4 + a5SOC 5 + a6SOC 6 + a7SOC 7 + a8SOC 8 + a9SOC 9 + a 10 SOC 10 +

[0097] a 11 SOC 11 + a 12 SOC 12, where a0, a1, …, a 12 are coefficients to be determined, and the OCV-SOC curve is obtained;

[0098] 2.4) Repeat steps 2.1 to 2.3 from 10 °C to -40 °C (at 5 °C intervals) until all OCV-SOC curves (a total of 9 curves) in the entire low-temperature range are fitted.

[0099] Then, use spline interpolation to obtain OCV values at 1 °C temperature intervals for different SOC values, and use polynomial fitting to obtain the OCV-SOC curves in each low-temperature range. The specific method is as follows:

[0100] 2.5) At 5 °C intervals, respectively extract the OCV values at different temperatures corresponding to the same SOC value (from 0 to 1, at 0.05 intervals);

[0101] 2.6) Use cubic spline interpolation to obtain all corresponding OCV values in the low-temperature range. The steps of spline interpolation are as follows: First, for the given data point sequence (T0, OCV0), (T1, OCV1), (T2, OCV2), …, (T 10 , OCV 10 ), where T X is the measured temperature, and OCV X is the open-circuit voltage corresponding to the SOC and temperature at this time (i.e., the average OCV obtained in step 2.2), calculate the distance h i = T i+1 - T i (i = 0, 1, …, 9). The parameters to be calculated are as follows:

[0102]

[0103] where μ i , λ i , d i are the relevant coefficients for constructing the equations,

[0104]

[0105] Solve for M i , and for each pair of adjacent data points (T i , OCV i ) and (T i+1 , OCV i+1 ) (i = 0, 1, …, 9), perform the following operations:

[0106]

[0107] where T i1 , T i2 , T i3 , Ti4 are the temperatures of the 4 inserted points, and they are evenly distributed in the interval [T i , T i+1 , and calculate the temperature intervals (1 °C) at the same SOC value:

[0108]

[0109] where In this way, all OCV values from -10 °C to 40 °C (at an interval of 1 °C) at the same SOC value can be calculated. Roll the SOC value and continue to use this method to obtain all groups of data (including SOC, T, OCV) with an interval of 0.05 from SOC = 0 to 1;

[0110] 2.7) Repeat steps 2.5 - 2.6 to obtain all OCV values corresponding to different SOCs (at an interval of 0.05) at different temperatures (at an interval of 1 °C);

[0111] 2.8) Using the data points obtained in step 2.7 above (each point contains 3 dimensions at this time), extract the data points at the same temperature T (the abscissa is SOC and the ordinate is OCV), and use a 12th - order polynomial fitting to obtain the OCV - SOC curves at an interval of 1 °C in each low - temperature range.

[0112] Fit the C - T curve through polynomial fitting to obtain the discharge capacity C at each temperature. The specific method is as follows:

[0113] Using the temperature T (at an interval of 5 °C) in the first step and the corresponding battery discharge capacity C obtained (by ampere - hour integration method), a total of 9 points. Taking the temperature T as the abscissa and the discharge capacity C as the ordinate, use a 3rd - order polynomial fitting to obtain the C - T curve, and use the 3rd - order polynomial function to estimate the discharge capacity C corresponding to all temperatures in the low - temperature range (at an interval of 1 °C).

[0114] In the third step, first establish a second - order RC equivalent circuit model according to the battery (this circuit includes a voltage source representing the open - circuit voltage U ocv , a series resistance R0, and two parallel RC networks, namely R1, C1 and R2, C2, where U is the terminal voltage, and the circuit diagram can be seen in the attached drawings of the specification):

[0115]

[0116] where U ocvis the open-circuit voltage (i.e., the OCV obtained in the second step, obtained from the SOC and OCV-SOC curve at the previous moment), R0 is the ohmic internal resistance, R1 and C1 are the electrochemical polarization resistance and capacitance respectively, R2 and C2 are the concentration polarization resistance and capacitance respectively, U1 and U2 are the electrochemical polarization voltage (the voltage across R1) and concentration polarization voltage (the voltage across R2) respectively, U is the voltage across the circuit (i.e., the measured voltage during the actual operation of the battery), and I is the current across the circuit (i.e., the measured current during the actual operation of the battery).

[0117] Then, the adaptive recursive least squares method is used for parameter identification to identify R0, R1, C1, R2, and C2 in the equivalent circuit. Specifically:

[0118] One is initialization, setting the initial parameter estimation values and the initial covariance matrix P0. The initial covariance matrix P0 generally takes a relatively large 5×5 matrix as large as possible.

[0119] Two is to calculate the predicted output y(k) according to the current parameter estimation values and the input :

[0120]

[0121] where

[0122] Three is to calculate the gain matrix K k , update the parameter estimation values and update the covariance matrix P k :

[0123]

[0124] where y(k) = U ocv (k) - U(k).

[0125] Four is to update the dynamic genetic factor λ k :

[0126]

[0127] where α is an adjustable parameter less than 1 and close to 1; γ is a positive adaptability coefficient.

[0128] Five is to perform parameter identification and convert to R0, R1, R2, C1, and C2:

[0129]

[0130] where T is the sampling period, a = τ1τ2, b = τ1 + τ2. Repeat steps two to five above until all data are identified. Then, take the mean values of R0, R1, R2, C1, and C2 respectively to obtain the input-adaptive extended Kalman filter algorithm module.

[0131] Finally, use the adaptive extended Kalman filter algorithm to calculate the real-time SOC value, specifically:

[0132] One is to establish the state equation and observation equation according to the equivalent circuit model:

[0133]

[0134] U k =-U 1,k -U 2,k +U OCV,k -I k R0 (14)

[0135] where η is the charge-discharge efficiency (set to 1), Q is the actual discharge capacity of the battery (i.e., the discharge capacity C in step 1), U 1,k is the voltage across R1 at time K, U 2,k is the voltage across R2 at time K, τ1 and τ2 have the same equivalent relationship as in equation (12), and Ts is the sampling period;

[0136] Two is to initialize the state vector Initialize the covariance matrix P0 (generally take a 3×3 positive definite matrix), set the process noise covariance matrix Q0 and the measurement noise covariance R0, and initialize the Jacobian matrix H0 = [-1 -1 0]. These matrices can be initially estimated and set according to the noise characteristics of the system.

[0137] Three is to perform prediction. According to the state equation, perform the prior state prediction for the next moment Use the prior estimate value and the OCV-SOC curve corresponding to the temperature to calculate the open-circuit voltage U OCV,k , through, update the observation matrix H k =[-1 -1U OCV,k , use the observation equation to estimate the terminal voltage Perform the prior state covariance prediction:

[0138]

[0139] Four is the update step. Calculate the Kalman gain:

[0140]

[0141] Update the posterior state value:

[0142]

[0143] Among them, U k is the measured terminal voltage value, and the error covariance matrix is updated as follows:

[0144] P k =(I е -K k H k )P k / k-1 (19)

[0145] Among them, I е is the identity matrix, adaptively adjust the measurement noise covariance R k and the process noise covariance matrix Q k :

[0146]

[0147] Among them α and β represent adjustment factors, and e k is the error between the measured terminal voltage and the estimated terminal voltage.

[0148] Loop the above steps 3 to 4 until all data is predicted, and output the state vectors at all times Extract all SOC values.

[0149] It should be noted that the temperature interval of the low-current constant-current discharge experiment is not necessarily 5°C, it can be 10°C, but the larger the temperature interval, the greater the error. The present invention is directed to lithium iron phosphate batteries, but for some batteries, due to manufacturing reasons, they cannot be discharged at low temperatures such as -40°C. Then, the lowest working temperature is taken as the lower limit, and the working steps are the same.

[0150] Low-current constant-current charge and discharge experiment: To ensure the accuracy of the data, repeated experiments can be appropriately carried out, but avoid excessive repetition, which will affect the battery life and thus the SOC estimation result.

[0151] SOC estimation: Since the temperature of the lithium battery will increase to a certain extent and be unstable during operation, when inputting the temperature, please first round the temperature data to an integer to represent the temperature situation around this temperature (±0.5°C) to achieve full coverage of low temperatures.

[0152] Embodiment

[0153] To verify the effectiveness of the proposed solution of the present invention, experiments are carried out according to the experimental steps in the specific implementation. After obtaining the results, output the state vectors at all times Extract all SOC values. As Figure 6 shown, this figure shows the results under the DST working condition when the temperature T = -2°C. From Figure 6As can be seen from the left - hand figure error analysis, the SOC estimation error after temperature correction (red curve) fluctuates within a small range around 0, and the error is controlled within ±0.05, effectively suppressing the error growth; while the error without temperature correction (blue curve) continuously decreases over time, and the error reaches about - 0.2, indicating that the error will accumulate significantly over time without correction. Observe Figure 6 In the right - hand figure, the SOC estimated value after temperature correction (red) almost coincides with the theoretical value (black), with high accuracy; while the estimated value without temperature correction (blue) deviates significantly from the theoretical value, and the gap further expands in the later stage. Overall, temperature correction improves the SOC estimation accuracy by compensating for the temperature effect, controls the error accumulation, and makes the result closer to the theoretical true value; without temperature correction, the error deteriorates over time, resulting in the estimated result deviating from the actual situation, fully demonstrating the necessity of temperature correction in battery SOC estimation.

[0154] The parameter identification and SOC estimation in the above - mentioned embodiments can be combined arbitrarily. For the sake of simplicity of description, not all possible combinations of parameter identification and SOC estimation in the above - mentioned embodiments are described. However, as long as the combination of these technical features does not conflict, the effectiveness of the present invention can be verified.

[0155] It should be noted that for those of ordinary skill in the art, without departing from the concept of the present application, several modifications and improvements can still be made, and these all fall within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the appended claims.

Claims

1. A method for estimating SOC of a lithium iron phosphate battery considering the influence of low temperature, characterized in that: The steps include: Step 1, through a low current constant current charge and discharge experiment, obtain the voltage and current data of the temperature gradient in the low temperature range, and calculate the battery remaining power value SOC and discharge capacity C at each sampling point; Step 2: Use the interpolation method to synchronize the SOC sampling intervals at the temperature of each gradient, obtain new data points by taking the average of the charge and discharge curves, and use polynomial fitting to obtain the OCV-SOC curve in the low temperature range; use the spline interpolation method to obtain the OCV value with a temperature interval of 1°C under different SOCs, and use polynomial fitting to obtain the OCV-SOC curve in each low temperature range; use the polynomial fitting to obtain the CT curve and obtain the discharge capacity C at each temperature; Step 3: Establish the battery equivalent circuit model and state equation, use the adaptive recursive least squares method to perform online parameter identification, and then calculate the real-time SOC value based on the adaptive extended Kalman filter algorithm.

2. The SOC estimation method for low temperature correction of lithium iron phosphate battery according to claim 1, characterized in that: Step 1: Through low-current constant-current charge and discharge experiments, obtain the voltage and current data of the low-temperature range, and calculate the battery remaining power value SOC and discharge capacity C at each sampling point. The specific method is: 1) Discharge at a constant current C until the battery is fully discharged; 2) Adjust the temperature of the thermostat and place the lithium battery in the thermostat for 30 minutes; 3) Charge the battery according to the standard charging protocol until the current drops to 0 in the constant voltage mode, and collect the voltage U and current I during the charging process; 4) Discharge at a constant rate of C / 20 until the battery is fully discharged, and collect voltage and current data during the discharge process; 5) Collect voltage (U) and current (I) data in the low temperature range at intervals of 5°C from 10°C to -40°C; 6) Calculate the battery remaining charge (SOC) value and discharge capacity C at each sampling point.

3. The SOC estimation method for low temperature correction of lithium iron phosphate battery according to claim 1, characterized in that: Step 2: Use the interpolation method to synchronize the SOC sampling intervals at each gradient temperature, obtain new data points by taking the average of the charge and discharge curves, and use polynomial fitting to obtain the OCV-SOC curve in the low temperature range; use the spline interpolation method to obtain the OCV value with a temperature interval of 1°C under different SOCs, and use polynomial fitting to obtain the OCV-SOC curve in each low temperature range; use the polynomial fitting to obtain the CT curve and obtain the discharge capacity C at each temperature. The specific method is as follows: 2.1) Use the interpolation method to synchronize the SOC sampling intervals at each temperature level, obtain new data points by taking the average of the charge and discharge curves, and use polynomial fitting to obtain the OCV-SOC curve in the low temperature range; 2.1.1) Based on the battery remaining capacity value SOC and the voltage U in the charge and discharge experiment, use the interpolation method to evenly divide all the charging and discharging data points (SOC, U) of the gradient temperature into 1000 points, and calculate the value of the interpolation point: First calculate the slope Then calculate the point OCV required for interpolation C =OCV0+k(SOC-SOC0), where OCV1 is the voltage corresponding to the next SOC point, OCV0 is the voltage corresponding to the previous SOC point, SOC1 and SOC0 are two points adjacent to the required interpolation point, OCV C is the interpolated voltage obtained in the middle, and SOC is the point corresponding to the interpolation in the middle; 2.1.2) Considering the hysteresis characteristics of the open circuit voltage, the average value of the charge interpolation voltage and the discharge interpolation voltage is taken to obtain the corresponding open circuit voltage average value OCV = (OCV C +OCV D ) / 2, where OCV C Indicates the charging interpolation voltage, OCV D represents the discharge interpolation voltage; 2.1.3) Using a 12th-order polynomial fitting, the OCV-SOC curve corresponding to the gradient temperature is obtained: OCV = a0 + a1SOC + a2SOC 2 +a3SOC 3 +a4xSOC 4 +a5SOC 5 +a6SOC 6 +a7SOC 7 +a8SOC 8 +a9SOC 9 +a 10 SOC 10 +a 11 SOC 11 +a 12 SOC 12 , where a0, a1, …, a 12 is the coefficient to be determined; 2.1.4) From 10°C to -40°C, repeat steps 2.1 to 2.3 at intervals of 5°C until all OCV-SOC curves in the low temperature range are fitted; 2.2) Use the spline interpolation method to obtain the OCV value at a temperature interval of 1°C under different SOCs, and use polynomial fitting to obtain the OCV-SOC curve in each low temperature range. The specific method is as follows: 2.2.1) Take the OCV values ​​of different temperatures corresponding to the same SOC value at intervals of 5°C. The SOC value range is from 0 to 1, and the sampling interval is 0.05; 2.2.2) Use cubic spline interpolation to obtain all OCV values ​​corresponding to the low temperature range. The steps of spline interpolation are as follows: First, for the given data point sequence (T0, OCV0), (t1, OCV1), (T2, OCV2), …, (T 10 ,OCV 10 ), T X is the measured temperature, OCV X For the open circuit voltage corresponding to SOC and temperature at this time, calculate the distance h between adjacent nodes i =T i+1 -T i (i=0,1,…,9), the parameters to be calculated are as follows: where μ i , i ,d i To construct the correlation coefficient of the equation system, Solve M i , for each pair of adjacent data points (T i ,OCV i ) and (T i+1 ,OCV i+1 )(i=0,1,…,9) perform the following operations: Where T i1 ,T i2 ,T i3 ,T i4 are the temperatures of the four inserted points, and they are evenly distributed in the interval [T i ,T i+1 ], calculate all OCV values ​​at each temperature interval under the same SOC value: in After calculating all OCV values ​​at the same SOC value, from -10℃ to 40℃, with an interval of 1℃, the SOC value is rolled over to obtain the SOC, T, and OCV combinations of all groups with an SOC interval of 0.05 from 0 to 1; 2.2.3) Take the data points at the same temperature T, with SOC as the horizontal axis and OCV as the vertical axis, and use a 12th-order polynomial fit to obtain the OCV-SOC curve at intervals of 1°C in each low temperature range; 2.3) The CT curve is fitted by polynomials to obtain the discharge capacity C at each temperature. The specific method is: With temperature T as the horizontal axis and discharge capacity C as the vertical axis, a third-order polynomial is used to fit the CT curve. The discharge capacity C corresponding to all temperatures within an interval of 1°C in the low temperature range is estimated using the third-order polynomial function.

4. The SOC estimation method for low temperature correction of lithium iron phosphate battery according to claim 1, characterized in that: Step 3: Establish the battery equivalent circuit model and state equation, use the adaptive recursive least squares method for online parameter identification, and then calculate the real-time SOC value based on the adaptive extended Kalman filter algorithm. The specific method is as follows: 3.1) Establish a second-order RC equivalent circuit model based on the battery, including a voltage source representing the open circuit voltage U of the battery ocv , a series resistor R0 and two parallel RC networks, namely R1, C1 and R2, C2, are expressed as: Among them U ocv is the open circuit voltage, obtained from the SOC and OCV-SOC curves at the previous moment, R0 is the ohmic internal resistance, R1 and C1 are the electrochemical polarization resistance and capacitance respectively, R2 and C2 are the concentration polarization resistance and capacitance respectively, U1 and U2 are the electrochemical polarization voltage and concentration polarization voltage respectively, U is the voltage across the circuit, i.e., the measured voltage of the actual operation of the battery, and I is the current across the circuit, i.e., the measured current of the actual operation of the battery; 3.2) Use the adaptive recursive least squares method to perform parameter identification and identify R0, R1, C1, R2, and C2 in the equivalent circuit, specifically: 3.2.1) Initialization, setting initial parameter estimates and the initial covariance matrix P0; 3.2.2) Based on the current parameter estimates and input Calculate the predicted output y(k): in 3.2.3) Calculate the gain matrix K k , update the parameter estimates And update the covariance matrix P k : where y(k) = U ocv (k) - U(k); 3.2.4) Update dynamic genetic factor λ k : in α is an adjustable parameter less than 1 and close to 1; γ is a positive adaptability coefficient; 3.2.5) Perform parameter identification and Convert to R0, R1, R2, C1, C2: Where t is the sampling period, a=τ1τ2, b=τ1+τ2, and repeat 3.2.2) to 3.2.5) until all data are identified; then, the average of R0, R1, R2, C1, and C2 is taken to obtain the input adaptive extended Kalman filter algorithm module; 3.3) Use the adaptive extended Kalman filter algorithm to calculate the real-time SOC value, specifically: 3.3.1) Establish the state equation and observation equation based on the battery equivalent circuit model: IN k =-U 1,k -IN 2,k +U OCV,k -AND k R0 (14) Where η is the charge and discharge efficiency, Q is the actual discharge capacity of the battery, and U 1,k is the voltage across R1 at time K, U 2,k is the voltage across R2 at time K, τ1 and τ2 have the same equivalent relationship as in equation (12), and Ts is the sampling period; 3.3.2) Initialize the state vector Initialize the covariance matrix P0, set the process noise covariance matrix Q0 and the measurement noise covariance R0, and initialize the Jacobian matrix H0 = -1-1 0], these matrices are preliminarily estimated and set according to the noise characteristics of the system; 3.3.3) Make a prediction and make a priori state prediction for the next moment based on the state equation Calculate the open circuit voltage U using the prior estimate and the OCV-SOC curve at the corresponding temperature OCV,k , update the observation matrix H by k =[-1 -1 U OCV,k ], using the observation equation to estimate the terminal voltage Make a priori state covariance prediction: 3.3.4) Update step, calculate the Kalman gain: Posteriori state value update: Among them U k is the measured terminal voltage value, and the error covariance matrix is ​​updated: P k =( е -K k H k )P k / k-1 (19) Among them I е is the unit matrix, adaptively adjust the measurement noise covariance R k Process noise covariance matrix Q k : in α and β represent adjustment factors, e k is the error between the measured terminal voltage and the estimated terminal voltage. Repeat steps 3.3.3) to 3.3.4) until all data are predicted and the state vectors at all times are output. Get all SOC values.

5. A lithium iron phosphate battery SOC estimation system considering the influence of low temperature, characterized in that: The method for estimating the SOC of a lithium iron phosphate battery considering the influence of low temperature as described in any one of claims 1 to 4 is implemented to realize the SOC estimation of a lithium iron phosphate battery considering the influence of low temperature, and steps 1 to 3 are respectively performed in three modules.

6. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the method for estimating the SOC of a lithium iron phosphate battery taking into account the influence of low temperature as described in any one of claims 1 to 4 is implemented to achieve SOC estimation of a lithium iron phosphate battery taking into account the influence of low temperature.

7. A computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the method for estimating the SOC of a lithium iron phosphate battery considering the influence of low temperature according to any one of claims 1 to 4 is implemented to achieve SOC estimation of a lithium iron phosphate battery considering the influence of low temperature.

Images