Method and system for estimating SOC (State of Charge) value of battery
By combining Coulomb counting method and voltage model in BMS and fusion using Kalman filtering algorithm, the problem of insufficient battery SOC estimation accuracy in low temperature environments is solved, and higher battery SOC estimation accuracy and electric vehicle performance are achieved.
Patent Information
- Application Number
- CN202510309766.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-03-17
- Publication Date
- 2025-07-01
Smart Images
Figure CN120233234A_ABST
Abstract
Description
Technical Field
[0001] The present disclosure belongs to the technical field of battery SOC estimation, and particularly relates to a method and system for estimating the battery SOC value. Background Art
[0002] In the development process of electric vehicles, the Battery Management System (BMS) is one of the key technologies;
[0003] The existing BMS has insufficient accuracy in estimating the battery SOC in a low-temperature environment, which affects the performance and user experience of electric vehicles; in a low-temperature environment, the discharge capacity of the battery and the accuracy of the battery SOC will be significantly reduced, resulting in a decrease in the utilization rate and service life of the battery;
[0004] For example, in an environment of minus 20 degrees Celsius, the capacity of a lithium battery may be only about 50% of that at room temperature, which has a significant impact on the endurance and performance of electric vehicles;
[0005] Therefore, a solution that can improve the accuracy of battery SOC estimation under low-temperature conditions is needed. Summary of the Invention
[0006] To solve the above problems, the present disclosure provides a method and system for estimating the battery SOC value, which uses the Kalman filter algorithm to fuse the predicted battery SOC value calculated by the Coulomb counting method and the measured battery SOC value calculated by the voltage model, and can obtain a more accurate estimated value of the battery SOC in a low-temperature state.
[0007] The following are the technical details of the present disclosure:
[0008] A method for estimating the battery SOC value, characterized by comprising:
[0009] Calculating the battery SOC value by using the Coulomb counting method to obtain the predicted battery SOC value;
[0010] Using the relationship between the open-circuit voltage of the battery and the battery SOC value to construct a voltage model, where the voltage model is a multi-dimensional relationship curve between the open-circuit voltage, temperature, discharge rate, battery internal resistance, and battery SOC value; according to the measured values of the open-circuit voltage, temperature, discharge rate, and battery internal resistance of the battery, obtaining the measured battery SOC value through the multi-dimensional relationship curve;
[0011] Using the Kalman filtering algorithm, fuse the measured value and predicted value of the battery SOC to obtain the final battery SOC value; during the fusion process, adjust the process noise covariance matrix Q in the Kalman filtering algorithm based on the battery temperature; adjust the measurement noise covariance matrix R in the Kalman filtering algorithm based on the relationship between the battery temperature and the battery discharge rate.
[0012] Further,
[0013] Calculate the battery SOC value using the Coulomb counting method;
[0014] The calculation formula of the Coulomb counting method is:
[0015]
[0016] where, SOC(t0) is the battery SOC value at the initial time t0, I(t) is the measured current value at time t; C nom is the nominal capacity of the battery; SOC(t) is the battery SOC value at time t.
[0017] Further,
[0018] Using the Kalman filtering algorithm, fuse the measured value and predicted value of the battery SOC to obtain the final battery SOC value; including:
[0019] Construct a state transition equation according to the calculation formula of the Coulomb counting method, and the state transition equation is used to calculate the battery SOC value at the next moment based on the battery SOC value at the previous moment, the current at the current moment, and the nominal capacity of the battery, as the predicted value of the battery SOC;
[0020] Construct a measurement equation based on the voltage model, and the measurement equation is a functional relationship between the battery SOC value at the current moment, the current temperature, the current current, the measurement noise, and the voltage value at the current moment;
[0021] Adjust the process noise covariance matrix Q according to the pre-established mapping relationship between the temperature and the process noise covariance; adjust the measurement noise covariance matrix R according to the relationship between the temperature and the battery discharge rate;
[0022] Calculate the covariance matrix at the current moment based on the process noise covariance matrix Q;
[0023] Determine the observation matrix according to the measurement equation; calculate the Kalman gain using the covariance matrix at the current moment, the measurement noise covariance matrix R, and the observation matrix;
[0024] Calculate the battery SOC correction value using the Kalman gain, the predicted value of the battery SOC, the measured value of the battery SOC, and the observation matrix;
[0025] Update the covariance matrix. After each update of the covariance matrix, if the algorithm termination condition is satisfied, end the update. At this time, the corrected value of the battery SOC obtained is the final battery SOC value.
[0026] Furthermore,
[0027] The state transition equation is:
[0028]
[0029] The measurement equation is:
[0030] v k = f(SOC k , T k , I k ) + ω k
[0031] where ω k is the measurement noise at time k, T k is the temperature at time k; SOC k is the battery SOC value at time k; I k is the current at time k; SOC k+1 is the battery SOC value at time k + 1;
[0032] The corrected value of the battery SOC is calculated using the Kalman gain, the predicted value of the battery SOC, the measured value of the battery SOC, and the observation matrix; the formula is:
[0033] SOC k+1|k+1 = SOC k+1|k + K k (V k - H k SOC k+1|k )
[0034] where V k is the measured voltage value at time k, H k is the observation matrix at time k; SOC k+1|k is the predicted SOC value at time k + 1 obtained based on time k; SOC k+1|k+1 is the corrected value of the battery SOC at time k + 1.
[0035] Furthermore,
[0036] The covariance matrix is updated; the formula is:
[0037] P k+1|k+1 = (I - K k H k )P k+1|k
[0038] Among them, K k is the current Kalman gain, H k is the observation matrix at time k, P k+1|k is the covariance matrix at time k+1 obtained based on time k; I is the identity matrix; P k+1|k+1 is the updated covariance matrix.
[0039] Furthermore,
[0040] The establishment of the three-dimensional characteristic curves of voltage, temperature, and battery SOC includes:
[0041] Set fitting parameters. The independent variables of the fitting parameters include: temperature, discharge rate, battery SOC value, and battery internal resistance; the dependent variable of the independent variables of the fitting parameters is the open-circuit voltage;
[0042] Sample the open-circuit voltage of the battery under different temperatures, different discharge rates, different internal resistances, and different battery SOC values to form a data set;
[0043] Use the data set for multiple linear regression fitting to obtain the first voltage model;
[0044] Build a non-linear fitting model based on the kernel function, and use the data set to train the support vector machine to obtain the second voltage model;
[0045] Test the first voltage model and the second voltage model through cross-validation and experimental data, and select the model with the smallest error as the final voltage model.
[0046] Furthermore,
[0047] The first voltage model is:
[0048] OCV = a0 + a1T + a2C rate + a3SOC + a i R + ε o
[0049] Among them, a i is the i-th fitting coefficient, ε o is the error term; C rate is the discharge rate; T is the temperature; R is the resistance; SOC is the state of charge of the battery;
[0050] The second voltage model is:
[0051]
[0052] Among them, K(x i , x) is the kernel function, α ia and b are fitting parameters; OCV is the open circuit voltage; N represents the number of samples used to build the model.
[0053] Furthermore,
[0054] It further includes: when calculating the battery SOC value in the low-temperature state using the Coulomb counting method, compensating the current measurement value in the Coulomb counting method to obtain a current compensation value; calculating the battery SOC value in the low-temperature state based on the current compensation value;
[0055] Compensating the open circuit voltage measurement value in the voltage model to obtain a voltage compensation value; obtaining the battery SOC measurement value through the multi-dimensional relationship curve based on the voltage compensation value, temperature, discharge rate, and battery internal resistance measurement value;
[0056] The voltage compensation value is:
[0057] V comp = V meas + f(T, R int , I)
[0058] where, V meas is the initial voltage measurement value, T is the temperature, R int is the battery internal resistance, I is the current; V comp is the voltage compensation value; f is the compensation function;
[0059] The current compensation value is:
[0060] I comp = I meas × (1 + k(T))
[0061] where, I meas is the initial current measurement value, I comp is the current compensation value; k(T) is a correction factor related to the temperature T.
[0062] An estimation system for battery SOC, characterized by including:
[0063] A first prediction module, used to calculate the battery SOC value using the Coulomb counting method to obtain a battery SOC predicted value;
[0064] A second prediction module, used to construct a voltage model using the relationship between the open circuit voltage of the battery and the battery SOC value, where the voltage model is a multi-dimensional relationship curve between the open circuit voltage, temperature, discharge rate, battery internal resistance, and battery SOC value; obtaining the battery SOC measurement value through the multi-dimensional relationship curve according to the open circuit voltage, temperature, discharge rate, and battery internal resistance measurement values of the battery;
[0065] A fusion module is used to fuse the measured value and predicted value of the battery SOC by using the Kalman filtering algorithm to obtain the final battery SOC value. During the fusion process, the process noise covariance matrix Q in the Kalman filtering algorithm is adjusted based on the battery temperature, and the measurement noise covariance matrix R in the Kalman filtering algorithm is adjusted based on the relationship between the battery temperature and the battery discharge rate.
[0066] Furthermore,
[0067] Specifically, the fusion module is used for:
[0068] Construct a state transition equation according to the calculation formula of Coulomb counting method. The state transition equation is used to calculate the battery SOC value at the next moment based on the battery SOC value at the previous moment, the current at the current moment, and the nominal capacity of the battery, as the predicted value of the battery SOC.
[0069] Construct a measurement equation based on the voltage model. The measurement equation is a functional relationship between the battery SOC value at the current moment, the current temperature, the current current, the measurement noise, and the voltage value at the current moment.
[0070] Adjust the process noise covariance matrix Q according to the pre-established mapping relationship between temperature and process noise covariance, and adjust the measurement noise covariance matrix R according to the relationship between temperature and battery discharge rate.
[0071] Calculate the covariance matrix at the current moment based on the process noise covariance matrix Q.
[0072] Determine the observation matrix according to the measurement equation, and calculate the Kalman gain by using the covariance matrix at the current moment, the measurement noise covariance matrix R, and the observation matrix.
[0073] Calculate the battery SOC correction value by using the Kalman gain, the predicted value of the battery SOC, the measured value of the battery SOC, and the observation matrix.
[0074] Update the covariance matrix. After each update of the covariance matrix, if the algorithm end condition is satisfied, the update is ended. At this time, the obtained battery SOC correction value is the final battery SOC value.
[0075] Compared with the prior art, the present disclosure has the following advantages:
[0076] The present disclosure uses the Coulomb counting method to calculate the change in battery SOC by integrating the charging and discharging current of the battery, which can track the change in power in real time and provide a preliminary battery SOC prediction value. However, the Coulomb counting method has cumulative errors; while the voltage model uses the multi-dimensional relationship curve between the open-circuit voltage and the battery SOC, considering factors such as temperature, discharge rate, and battery internal resistance, which can more accurately describe the relationship between the open-circuit voltage and the battery SOC at low temperatures and obtain a relatively accurate battery SOC measurement value. The combination of the two can complement each other's advantages and make up for their respective deficiencies;
[0077] Taking the Kalman filter algorithm as the optimal estimation method, it can perform weighted fusion on the battery SOC prediction value and the measurement value according to the system dynamic characteristics and the measurement noise statistical characteristics, and adapt to the changes in the battery system under low-temperature environments by adjusting the process noise covariance matrix and the measurement noise covariance matrix, automatically adjusting the weights, so that the final battery SOC estimation value is more accurate.
[0078] Other features and advantages of the present disclosure will be described in the subsequent specification, and, in part, will be obvious from the specification, or will be understood by implementing the present disclosure. The objectives and other advantages of the present disclosure can be realized and obtained by the structures pointed out in the specification, the claims, and the drawings. Brief Description of the Drawings
[0079] In order to more clearly illustrate the technical solutions in the embodiments of the present disclosure or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present disclosure. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.
[0080] Figure 1 Shows a schematic diagram of the method of the present invention. Detailed Description of the Embodiments
[0081] To make the objectives, technical solutions, and advantages of the embodiments of the present disclosure clearer, the following will clearly and completely describe the technical solutions in the embodiments of the present disclosure with reference to the drawings in the embodiments of the present disclosure. Obviously, the described embodiments are some, but not all, of the embodiments of the present disclosure. Based on the embodiments in the present disclosure, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the scope of protection of the present disclosure.
[0082] Figure 1 Shows a schematic diagram of the method according to the present invention. The specific implementation details of the present invention include:
[0083] Battery SOC estimation, compensating the battery SOC estimation results using multi-parameter comprehensive compensation, temperature change rate compensation, and temperature compensation algorithms.
[0084] Innovation point: An SOC estimation method combining Coulomb counting method and voltage model.
[0085] 1) Adjust the Coulomb counting method; 2) Adjust the voltage model; 3) Adaptively adjust the battery SOC estimation parameters through an improved Kalman filtering algorithm.
[0086] 1. Step 1: Calculate the battery SOC value using the Coulomb counting method.
[0087] Specifically, it includes: By measuring the charge and discharge current of the battery and combining the sampling time interval, calculate the cumulative charge of the battery. The formula is as follows:
[0088]
[0089] Among them, SOC(t0) is the battery SOC value at the initial time t0, I(t) is the current measurement value at time t; C nom is the nominal capacity of the battery; SOC(t) is the battery SOC value at time t.
[0090] To solve the cumulative error problem, dynamic calibration is carried out in combination with a regular full charge and full discharge calibration mechanism.
[0091] Among them, due to the measurement error in the low-temperature environment, I(t) can be compensated to obtain a more accurate current compensation value. The current compensation value is:
[0092] I comp = I meas ×(1 + k(T))
[0093] Among them, I meas is the initial current measurement value, I comp is the current compensation value; k(T) is the temperature-related correction factor.
[0094] 2. Step 2: Estimate the battery SOC value through the voltage model:
[0095] Using the relationship between the open-circuit voltage (OCV) of the battery and the battery SOC value, establish a three-dimensional characteristic curve of voltage-temperature-battery SOC based on historical experimental data, and estimate the battery SOC value through the look-up table method.
[0096] The present invention combines an improved voltage model and improves the accuracy of the voltage model through a multi-parameter fitting method. For example, considering multiple factors such as the discharge rate, temperature, and internal resistance of the battery, establish a more accurate voltage model through a multi-parameter fitting method to improve the accuracy of battery SOC estimation.
[0097] Specifically, multi-parameter fitting methods such as multiple linear regression and support vector machine (SVM) can be used to model the multi-dimensional data of the battery to obtain a more accurate voltage model. For example, by fitting the voltage data of the battery at different temperatures and different discharge rates, a multi-dimensional voltage model of the battery can be obtained, making the estimation of the battery SOC more accurate.
[0098] The specific steps of multiple linear regression fitting and SVM fitting methods are as follows:
[0099] The voltage model fitting is based on experimental data to establish a multi-dimensional relationship between the open circuit voltage (OCV), temperature (T), discharge rate (C-rate) of the battery and the battery SOC value. Specifically, it includes:
[0100] 1) Fitting parameters:
[0101] Independent variables: temperature (T), discharge rate (C-rate), battery SOC value, battery internal resistance (R).
[0102] Dependent variable: open circuit voltage (OCV).
[0103] 2) Data collection:
[0104] Under different temperatures (-30°C to 50°C), different discharge rates (0.1C to 3C) and different battery SOC values (0% to 100%), sample the open circuit voltage, internal resistance and other characteristics of the battery to form a multi-dimensional experimental data set.
[0105] 3) Fitting method:
[0106] Multiple linear regression:
[0107] OCV = a0 + a1T + a2C rate + a3SOC + a4R + ε o
[0108] where a i is the i-th fitting coefficient, and ε o is the error term; C rate is the discharge rate; T is the temperature; R is the internal resistance of the resistor; the battery SOC is the state of charge;
[0109] Support vector machine (SVM): Based on a kernel function (such as a Gaussian kernel), a non-linear fitting model is constructed, which is suitable for cases where the voltage characteristics are strongly non-linear:
[0110]
[0111] where K(x i , x) is the kernel function, and α ia and b are fitting parameters; OCV is the open-circuit voltage; N represents the number of samples used to build the model.
[0112] 4) Verification and optimization:
[0113] Test the fitting model through cross-validation and experimental data, and select the model with the smallest error as the final voltage model.
[0114] Among them, due to the measurement error in the low-temperature environment, the measured value of the open-circuit voltage can be compensated to obtain a more accurate voltage compensation value. Specifically, it includes:
[0115] Compensate and correct the measured battery voltage value according to the real-time changes of temperature, discharge rate, and battery internal resistance; the compensation algorithm is: by establishing a multi-dimensional characteristic curve of temperature and internal resistance and discharge rate, calculate the voltage compensation value:
[0116] V comp = V meas + f(T, R int , I)
[0117] Among them, V meas is the initial voltage measurement value, T is the temperature, R int is the internal resistance, I is the current; V comp is the voltage compensation value, and f is the compensation function.
[0118] 3. Step 3: Fusion estimation:
[0119] The Kalman filter algorithm is a commonly used correction method in existing battery SOC estimation methods, but its effect is limited in the low-temperature environment; the present invention proposes an improved Kalman filter algorithm to correct the battery SOC estimation.
[0120] Use the Kalman filter algorithm to fuse the estimation results of the Coulomb counting method and the voltage model to correct the battery SOC estimation error. The specific method is:
[0121] 1) Definition of the state space model:
[0122] State variable: The battery SOC value.
[0123] Measured value: Battery voltage, current.
[0124] State transition equation:
[0125] Among them, SOC k is the battery SOC value at the current moment; SOC k+1 is the battery SOC value at the next moment; I k is the current current; Δt is the unit time; SOC k+1is the battery SOC value at time k+1.
[0126] Measurement equation: v k = f(SOC k , T k , I k ) + ω k
[0127] where ω k is the measurement noise, T k is the current temperature; SOC k is the battery SOC value at the current time; I k is the current current. Where ω k is the measurement noise at time k, T k is the temperature at time k; SOC k is the battery SOC value at time k; I k is the current at time k.
[0128] 2) Temperature-related noise model:
[0129] Process noise: Dynamically adjust the process noise covariance matrix Q according to the temperature;
[0130] Measurement noise: Dynamically adjust the measurement noise covariance matrix R based on the temperature and discharge rate.
[0131] State prediction:
[0132] Predict the battery SOC value SOC k at the next time according to the state transition equation and the predicted value SOC k+1 of the battery SOC at the current time, and calculate the predicted covariance matrix;
[0133] The state transition equation is:
[0134] where I k is the current at the current time, C nom is the nominal capacity of the battery.
[0135] Covariance matrix:
[0136] where is the transpose of the state transition matrix; P k|k is the covariance matrix at the current time; Q k is the process noise covariance matrix.
[0137] Calculate the Kalman gain:
[0138] Calculate the Kalman gain K k :
[0139] Among them, H k is the observation matrix; R k is the measurement noise covariance matrix;
[0140] State update:
[0141] Obtain the current observation value (such as battery voltage, etc.), and combine it with the Kalman gain K k , the predicted SOC at the previous moment k+1|k , the observation matrix H k ; Update the battery SOC value at the current moment:
[0142] The update formula is:
[0143] SOC k+1|k+1 = SOC k+1|k + K k (V k - H k SOC k+1|k )
[0144] Among them, V k is the voltage measurement value at time k, and H k is the observation matrix at time k; SOC k+1|k is the predicted SOC value at time k + 1 obtained based on time k; SOC k+1|k+1 is the corrected battery SOC value at time k + 1.
[0145] Covariance update:
[0146] According to the Kalman gain K k , the observation matrix H k , the covariance matrix P at the current moment k+1|k ; Update the covariance matrix at the current moment;
[0147] The formula is: P k+1|k+1 = (I - K k H k )P k+1|k
[0148] Among them, K k is the Kalman gain at time k, H k is the observation matrix at time k, P k+1|k is the covariance matrix at time k + 1 obtained based on time k; I is the identity matrix; P k+1|k+1 is the updated covariance matrix.
[0149] After each covariance update, judge whether the error covariance P k+1|k+1 is less than the preset threshold; if it is less than the threshold, end the update, and the SOC k+1|k+1 value obtained at this time is the final battery SOC value.
[0150] 4) Optimization mechanism:
[0151] By introducing non - linear state variables (such as battery internal resistance) through the Extended Kalman Filter (EKF), the estimation accuracy is further improved.
[0152] Through the above steps, the Kalman filter can dynamically correct the battery SOC estimation error and adapt to the changes in battery performance under low - temperature environments.
[0153] Through this combined method, the deficiencies of a single algorithm are overcome, and the accuracy and stability of battery SOC estimation under low - temperature environments are improved.
[0154] The following are specific demonstration examples of the present invention:
[0155] Example 1: Application of the temperature compensation model
[0156] In specific implementation, the temperature compensation model of the present invention dynamically adjusts by collecting real - time temperature, voltage, and current data of the battery and using a multi - parameter comprehensive compensation strategy.
[0157] For example, in an environment of minus 10 degrees Celsius, by real - time monitoring the changes in battery voltage and current and combining historical data analysis, the compensation parameters are dynamically adjusted to improve the accuracy of battery SOC estimation. When the temperature changes rapidly, the compensation model can respond to the temperature change in real - time, quickly correct the battery voltage and current, and improve the accuracy of battery SOC estimation.
[0158] Example 2: Optimization of the battery SOC estimation method
[0159] The battery SOC estimation method of the present invention combines an improved Coulomb counting method and a voltage model, and realizes high - precision battery SOC estimation through multi - parameter fitting and adaptive adjustment.
[0160] For example, in a low - temperature environment, through a dynamic calibration mechanism, the cumulative error of the Coulomb counting method is dynamically adjusted according to real - time data to improve the accuracy of battery SOC estimation. At the same time, through a multi - parameter fitting method, an accurate voltage model is established, combined with an adaptive adjustment mechanism, and the estimation result is automatically corrected according to the actual usage of the battery to improve the accuracy of battery SOC estimation.
[0161] Example 3: Calibration of the Kalman filter algorithm
[0162] The Kalman filter algorithm of the present invention introduces a temperature - related noise model, and improves the accuracy and stability of battery SOC estimation by optimizing the update mechanism of the filter algorithm and adding joint filtering of multiple state variables.
[0163] For example, in a low-temperature environment, the actual state of the battery is more accurately reflected through a temperature-related noise model. By optimizing the update mechanism of the filtering algorithm, the convergence speed and stability of the filter are improved. By adding state variables such as the temperature and resistance of the battery, joint filtering processing is performed to improve the accuracy and stability of battery SOC estimation.
[0164] Based on the method of the present invention, the embodiments of the present disclosure also provide a system corresponding to the above method, which includes:
[0165] A first prediction module, configured to calculate the battery SOC value using the Coulomb counting method to obtain a predicted battery SOC value;
[0166] A second prediction module, configured to construct a voltage model using the relationship between the open-circuit voltage of the battery and the battery SOC value, where the voltage model is a multi-dimensional relationship curve between the open-circuit voltage, temperature, discharge rate, battery internal resistance, and battery SOC value; and obtain a measured battery SOC value through the multi-dimensional relationship curve according to the measured values of the open-circuit voltage, temperature, discharge rate, and battery internal resistance of the battery.
[0167] A fusion module, configured to fuse the measured battery SOC value and the predicted battery SOC value using the Kalman filtering algorithm to obtain a final battery SOC value; during the fusion process, adjust the process noise covariance matrix Q in the Kalman filtering algorithm based on the battery temperature; and adjust the measurement noise covariance matrix R in the Kalman filtering algorithm based on the relationship between the battery temperature and the battery discharge rate.
[0168] Although the present disclosure has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present disclosure.
Claims
1. A method for estimating a battery SOC value, characterized in that: include: The battery SOC value is calculated using the Coulomb counting method to obtain the battery SOC prediction value; Using the relationship between the open circuit voltage of the battery and the battery SOC value, a voltage model is constructed, wherein the voltage model is a multidimensional relationship curve between the open circuit voltage, temperature, discharge rate, battery internal resistance and battery SOC value; and according to the open circuit voltage, temperature, discharge rate and battery internal resistance measurement values of the battery, the battery SOC measurement value is obtained through the multidimensional relationship curve; Using a Kalman filter algorithm, the battery SOC measurement value and the battery SOC prediction value are integrated to obtain a final battery SOC value; During the fusion process, the process noise covariance matrix Q in the Kalman filter algorithm is adjusted based on the battery temperature; the measurement noise covariance matrix R in the Kalman filter algorithm is adjusted based on the relationship between the battery temperature and the battery discharge rate.
2. A method for estimating a battery SOC value according to claim 1, characterized in that: The battery SOC value is calculated by using the Coulomb counting method; The calculation formula of Coulomb counting method is: Where SOC(t0) is the battery SOC value at the initial time t0, I(t) is the current measurement value at time t; C nom is the nominal capacity of the battery; SOC(t) is the battery SOC value at time t.
3. A method for estimating a battery SOC value according to claim 2, characterized in that: The method of using the Kalman filter algorithm to fuse the battery SOC measurement value and the battery SOC prediction value to obtain a final battery SOC value includes: A state transfer equation is constructed according to the calculation formula of the Coulomb counting method, wherein the state transfer equation is used to calculate the battery SOC value at the next moment based on the battery SOC value at the previous moment, the current at the current moment, and the battery nominal capacity as a battery SOC prediction value; Constructing a measurement equation based on the voltage model, wherein the measurement equation is a functional relationship between the battery SOC value at the current moment, the current temperature, the current current, the measurement noise and the voltage value at the current moment; According to the pre-established mapping relationship between temperature and process noise covariance, the process noise covariance matrix Q is adjusted; according to the relationship between temperature and battery discharge rate, the measurement noise covariance matrix R is adjusted; Calculate the covariance matrix at the current moment based on the process noise covariance matrix Q; Determine the measurement matrix according to the measurement equation; calculate the Kalman gain using the covariance matrix at the current moment, the measurement noise covariance matrix R, and the measurement matrix; The battery SOC correction value is calculated using the Kalman gain, the battery SOC prediction value, the battery SOC measurement value, and the observation matrix; Update the covariance matrix. After each update of the covariance matrix, if the algorithm end condition is met, the update is terminated. The battery SOC correction value obtained at this time is the final battery SOC value.
4. A method for estimating a battery SOC value according to claim 3, characterized in that: The state transfer equation is: The measurement equation is: v k =f(SOC k ,T k ,I k )+ω k Among them, ω k is the measurement noise at time k, T k is the temperature at time k; SOC k is the battery SOC value at time k; I k is the current at time k; SOC k+1 is the battery SOC value at time k+1; Δt is the unit time; The battery SOC correction value is obtained by using the Kalman gain, the battery SOC prediction value, the battery SOC measurement value, and the observation matrix; the formula is: SOC k+1|k+1 =SOC k+1|k +K k (V k -H k SOC k+1|k ) Among them, V k is the voltage measurement value at time k, H k is the observation matrix at time k; SOC k+1|k is the SOC prediction value at time k+1 based on time k; SOC k+1|k+1 is the battery SOC correction value at time k+1.
5. The method for estimating a battery SOC value according to claim 3, characterized in that: The updated covariance matrix; formula is: P k+1|k+1 =(I-K k H k )P k+1|k Among them, K k is the Kalman gain at time k, H k is the observation matrix at time k, P k+1|k is the covariance matrix of k+1 moment obtained based on k moment; I is the identity matrix; P k+1|k+1 is the updated covariance matrix.
6. A method for estimating a battery SOC value according to claim 1, characterized in that: The three-dimensional characteristic curve of voltage, temperature and battery SOC is established; including: Set fitting parameters. The independent variables of fitting parameters include: temperature, discharge rate, battery SOC value, battery internal resistance; the independent variable dependent variables of fitting parameters include: open circuit voltage; The open circuit voltage of the battery is sampled at different temperatures, different discharge rates, different internal resistances, and different battery SOC values to form a data set; The data set is used to perform multivariate linear regression fitting to obtain the first voltage model; A nonlinear fitting model is constructed based on the kernel function, and a support vector machine is trained using the data set to obtain a second voltage model; The first voltage model and the second voltage model are tested through cross-validation and experimental data, and the model with the smallest error is selected as the final voltage model.
7. A method for estimating a battery SOC value according to claim 6, characterized in that: The first voltage model is: OCV=a0+a1T+a2C rate +a3SOC+a i R+ε o Among them, a i is the i-th fitting coefficient, ε o is the error term; C rate is the discharge rate; T is the temperature; R is the resistance; SOC is the state of charge of the battery; The second voltage model is: Among them, K(x i , x) is the kernel function, α i and b are fitting parameters; OCV is the open circuit voltage; N represents the number of samples used to build the model.
8. The method for estimating a battery SOC value according to claim 1, characterized in that: The method also includes: when calculating the battery SOC value at low temperature using the Coulomb counting method, compensating the current measurement value in the Coulomb counting method to obtain a current compensation value; and calculating the battery SOC value at low temperature based on the current compensation value; Compensating the open circuit voltage measurement value in the voltage model to obtain a voltage compensation value; obtaining the battery SOC measurement value through the multidimensional relationship curve based on the voltage compensation value and the temperature, discharge rate, and battery internal resistance measurement values; The voltage compensation value is: V comp =V meas +f(T,R int ,I) Among them, V meas is the initial voltage measurement value, T is the temperature, R int is the internal resistance of the battery, I is the current; V comp is the voltage compensation value; f is the compensation function; The current compensation value is: I como =I neas ×(1+k(T)) Among them, I meas is the initial current measurement value, I comp is the current compensation value; k(T) is the correction factor related to temperature T.
9. A battery SOC estimation system, characterized in that: include: The first prediction module is used to calculate the battery SOC value by using the Coulomb counting method to obtain the battery SOC prediction value; The second prediction module is used to construct a voltage model by using the relationship between the open circuit voltage of the battery and the battery SOC value, wherein the voltage model is a multidimensional relationship curve between the open circuit voltage, temperature, discharge rate, battery internal resistance and battery SOC value; and obtain the battery SOC measurement value through the multidimensional relationship curve according to the open circuit voltage, temperature, discharge rate and battery internal resistance measurement values of the battery; A fusion module, used to fuse the battery SOC measurement value and the battery SOC prediction value using a Kalman filter algorithm to obtain a final battery SOC value; During the fusion process, the process noise covariance matrix Q in the Kalman filter algorithm is adjusted based on the battery temperature; the measurement noise covariance matrix R in the Kalman filter algorithm is adjusted based on the relationship between the battery temperature and the battery discharge rate.
10. A battery SOC estimation system according to claim 9, characterized in that: The fusion module is specifically used for: A state transfer equation is constructed according to the calculation formula of the Coulomb counting method, wherein the state transfer equation is used to calculate the battery SOC value at the next moment based on the battery SOC value at the previous moment, the current at the current moment, and the battery nominal capacity as a battery SOC prediction value; Constructing a measurement equation based on the voltage model, wherein the measurement equation is a functional relationship between the battery SOC value at the current moment, the current temperature, the current current, the measurement noise and the voltage value at the current moment; According to the pre-established mapping relationship between temperature and process noise covariance, the process noise covariance matrix Q is adjusted; according to the relationship between temperature and battery discharge rate, the measurement noise covariance matrix R is adjusted; Calculate the covariance matrix at the current moment based on the process noise covariance matrix Q; Determine the observation matrix according to the measurement equation; Calculate the Kalman gain using the current covariance matrix, the measurement noise covariance matrix R, and the observation matrix; The battery SOC correction value is calculated using the Kalman gain, the battery SOC prediction value, the battery SOC measurement value, and the observation matrix; Update the covariance matrix. After each update of the covariance matrix, if the algorithm end condition is met, the update is terminated. The battery SOC correction value obtained at this time is the final battery SOC value.
Citation Information
Cited By
Virtual electric quantity display method, electronic equipment and storage medium
CN121955750A