Battery SOC estimation method, device, equipment and medium

Abstract

The embodiments of the present application provide a battery SOC estimation method, device, equipment and medium, which relate to the field of battery technology and can be applied to various new energy vehicle-related fields such as electric vehicles, hybrid vehicles, and hydrogen energy vehicles. The method includes: determining the real-time current corresponding to the target battery based on the sampling error of the battery sensor configured for the target battery and the sampling current obtained by the battery sensor; then inputting the real-time current into the ampere-hour integration algorithm to output the SOC estimate of the target battery; then inputting the predetermined SOC and the temperature-compensated open-circuit voltage conversion relationship and the SOC estimate into the adaptive extended Kalman filter algorithm to output the estimated value of the battery SOC. The method of the present application effectively solves the problem of insufficient accuracy of the estimated value of the SOC of the vehicle battery in the related art.

Description

Technical Field

[0001] The present application relates to the field of battery technology and can be applied to various new energy vehicle-related fields such as electric vehicles, hybrid vehicles, and hydrogen energy vehicles, and in particular to a battery SOC estimation method, device, equipment, and medium. Background Art

[0002] The SOC of a vehicle battery, that is, the battery's state of charge, and the accuracy of its numerical determination are crucial to the management of the vehicle battery. If the vehicle battery SOC is not determined accurately, there will be a difference between the actual SOC and the estimated SOC, affecting the battery's charge and discharge management, the use of the vehicle, etc., and thus causing safety hazards.

[0003] In the prior art, the SOC of a vehicle battery is usually determined by estimation. The commonly used estimation method is to obtain the remaining capacity of the battery based on the difference between the real-time actual capacity and the actual capacity of the battery, and use the ratio of the remaining capacity to the total capacity as the estimated value of the battery SOC. However, this algorithm does not take into account issues such as ambient temperature and battery aging, resulting in a constant difference between the estimated value and the actual value of the battery SOC. Summary of the Invention

[0004] The present application provides a battery SOC estimation method, device, equipment and medium to solve the problem of insufficient accuracy of the estimated value of the SOC of a vehicle battery in the related art.

[0005] In a first aspect, the present application provides a battery SOC estimation method, apparatus, device, and medium, including:

[0006] Determine the real-time current corresponding to the target battery based on the sampling error of the battery sensor configured for the target battery and the sampled current obtained by the battery sensor;

[0007] Input the real-time current into the ampere-hour integration algorithm and output the estimated SOC value of the target battery;

[0008] The predetermined SOC and the temperature-compensated open-circuit voltage conversion relationship and the SOC estimate are input into an adaptive extended Kalman filter algorithm to output an estimated value of the battery-corrected SOC, wherein the adaptive extended Kalman filter algorithm includes adaptively updated process noise and observation noise.

[0009] In one embodiment of the present disclosure, a real-time current corresponding to a target battery is determined based on a sampling error of a battery sensor configured for the target battery and a sampled current obtained by the battery sensor, including: obtaining preset sampling error parameters corresponding to the battery sensor, the preset sampling error parameters including a sampling error variance; using a difference between the sampling current and the sampling error variance as a correction value for the sampling current, and combining the correction value with the sampling current to obtain the real-time current.

[0010] In one embodiment of the present disclosure, the real-time current I obs for:

[0011] ,

[0012] Among them, I real It is used to represent the sampling current, and μ is used to represent the sampling error variance.

[0013] In one embodiment of the present disclosure, the ampere-hour integration algorithm is:

[0014] ,

[0015] Among them, SOC(t) represents the estimated SOC at time t, I(τ) is used to represent the real-time current at time τ, SOC0 is used to represent the initial SOC of the target battery, and the initial SOC is determined by the open circuit voltage conversion relationship based on temperature compensation, C normal It is used to indicate the rated capacity, and t represents the duration that the target battery is in the charging or discharging state.

[0016] In one embodiment of the present disclosure, an open circuit voltage conversion equation based on temperature compensation is determined in the following manner: the open circuit voltage corresponding to different ambient temperatures and different SOC values ​​is obtained, and the open circuit voltage, the corresponding ambient temperature and the corresponding SOC value are used as sample data; based on the distribution range of the SOC value, the sample data is divided into at least three interval groups; the sample data of each interval group is substituted into the fitting formula respectively to determine the fitting formula parameters corresponding to each interval group, wherein the fitting formula is used to express the calculation relationship between the open circuit voltage and the corresponding SOC value and the ambient temperature; the fitting formula corresponding to each interval group is combined to obtain a conversion equation, wherein the conversion equation includes at least three fitting formulas corresponding to different SOC distribution intervals.

[0017] In one embodiment of the present disclosure, the fitting formula is:

[0018] ,

[0019] Wherein, OCV is used to represent the open circuit voltage, T is used to represent the ambient temperature, and a0, a1, a2, a3, b1, and b2 are fitting formula parameters.

[0020] In one embodiment of the present disclosure, a predetermined SOC, an open-circuit voltage conversion formula based on temperature compensation, and a battery SOC estimate are input into an adaptive extended Kalman filter algorithm, and an estimated value of the battery SOC is output, including: determining the actual capacity of the target battery based on the preset temperature characteristics, cell type, battery health status and rated capacity of the target battery; inputting the actual capacity, the open-circuit voltage conversion formula based on temperature compensation, and the SOC estimate into the adaptive extended Kalman filter algorithm, and outputting the battery SOC estimate.

[0021] In one embodiment of the present disclosure, the actual capacity is expressed as:

[0022] ,

[0023] Among them, CAP represents the actual capacity of the target battery, F represents the preset temperature characteristic value, temp represents the real-time ambient temperature, celltype represents the type of battery cell, C normal It is used to indicate the rated capacity, and SOH indicates the battery state of health.

[0024] In one embodiment of the present disclosure, the actual capacity, the open-circuit voltage conversion relationship based on temperature compensation, and the SOC prediction value are input into an adaptive extended Kalman filter algorithm to output a battery SOC estimation value, including: taking SOC0 as the initial value and determining the corresponding initial parameters, wherein the initial parameters include the covariance matrix P0, the process noise Q, and the observation noise R, the initial parameters are all unit matrices, and the process noise and the observation noise are both covariance matrices; based on the initial values ​​and initial parameters, determining the SOC prediction value and prediction parameters after a set time, the prediction parameters include the prediction covariance matrix; based on the actual capacity and the conversion relationship, determining the dynamic parameters corresponding to the SOC prediction value, the dynamic parameters include the Jacobian matrix and the Kalman gain; based on the dynamic parameters, updating the SOC prediction value; based on the difference between the updated SOC prediction value and the SOC prediction value, adaptively updating the initial parameters; and using the updated SOC prediction value as the SOC estimation value.

[0025] In one embodiment of the present disclosure, the SOC prediction value is:

[0026] ,

[0027] Among them, SOC pre (k) represents the SOC prediction value corresponding to the k-th step, SOC(k-1) represents the SOC prediction value of the previous step corresponding to SOCpre(k), η represents the charge and discharge efficiency, I represents the real-time current, and Δt represents the time between adjacent steps;

[0028] The prediction parameters are:

[0029] ,

[0030] Among them, P pre (k) represents the prediction covariance matrix corresponding to the kth step length, P(k-1) represents P pre (k) is the prediction covariance matrix of the previous step corresponding to the k-th step, Q(k) represents the process noise covariance matrix corresponding to the k-th step;

[0031] The Jacobian matrix is:

[0032] ,

[0033] Among them, H k represents the Jacobian matrix corresponding to the k-th step size, and f represents the conversion relationship;

[0034] The Kalman gain is:

[0035] ,

[0036] Among them, K k represents the Kalman gain corresponding to the k-th step length, and R(k) represents the observation noise covariance matrix corresponding to the k-th step length;

[0037] The SOC prediction value is updated as:

[0038] ,

[0039] Among them, U up (k) represents the updated open circuit voltage corresponding to the kth step, U pre (k) represents the predicted open circuit voltage corresponding to the kth step, U ocv (k) represents the open circuit voltage determined based on the conversion relationship corresponding to the k-th step, SOC(k) represents the updated SOC prediction value corresponding to the k-th step, SOC pre (k) represents the SOC prediction value corresponding to the k-th step length;

[0040] The adaptive update of the initial parameters is expressed as:

[0041] ,

[0042] Among them, b is the forgetting factor, d is the weight parameter, k is the step size, Q(k-1) represents the process noise corresponding to the k-1th step size, K k ′ represents the derivative of the Kalman gain corresponding to the k-th step, epsilon represents the observation residual, and epsilon′ represents the derivative of the observation residual. The observation residual is used to represent the difference between the SOC estimate value obtained based on the open-circuit voltage in the k-th step and the SOC estimation value.

[0043] In a second aspect, an embodiment of the present disclosure provides a battery SOC estimation device, the battery SOC estimation device comprising:

[0044] A determination module, configured to determine a real-time current corresponding to a target battery based on a sampling error of a battery sensor configured for the target battery and a sampled current obtained by the battery sensor;

[0045] The estimation module is used to input the real-time current into the ampere-hour integration algorithm and output the estimated SOC value of the target battery;

[0046] The output module is used to input a predetermined SOC and a temperature-compensated open-circuit voltage conversion relationship and an estimated SOC value into an adaptive extended Kalman filter algorithm, and output an estimated value of the battery SOC, wherein the adaptive extended Kalman filter algorithm includes adaptively updated process noise and observation noise.

[0047] In a third aspect, an embodiment of the present disclosure further provides a control device, the control device comprising:

[0048] at least one processor;

[0049] and a memory communicatively coupled to the at least one processor;

[0050] The memory stores instructions that can be executed by at least one processor, and the instructions are executed by the at least one processor to enable the control device to perform the battery SOC estimation method as described in the first aspect of the present disclosure.

[0051] In a fourth aspect, an embodiment of the present disclosure further provides a computer-readable storage medium, which stores computer-executable instructions. When the computer-executable instructions are executed by a processor, they are used to implement the battery SOC estimation method as described in the first aspect of the present disclosure.

[0052] In a fifth aspect, an embodiment of the present disclosure further provides a computer program product, which includes computer execution instructions. When the computer execution instructions are executed by a processor, they are used to implement the battery SOC estimation method as described in the first aspect of the present disclosure.

[0053] The battery SOC estimation method, apparatus, device, and medium provided by the embodiments of the present disclosure determine the real-time current corresponding to the target battery by using the sampling error of the battery sensor configured based on the target battery and the sampled current obtained by the battery sensor. The real-time current is then input into an ampere-hour integration algorithm to output an estimated SOC value for the target battery. The predetermined SOC and the temperature-compensated open-circuit voltage conversion relationship and the estimated SOC value are then input into an adaptive extended Kalman filter algorithm to output an estimated battery SOC value. Thus, during the battery SOC estimation process, the sampling error of the sensor and the resulting error caused by the ambient temperature are fully considered. The accumulated error in the calculation process is then corrected by the Kalman filter algorithm, maximizing the accuracy of the battery SOC estimation value. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0055] Figure 1 A diagram of an application scenario of the battery SOC estimation method provided in an embodiment of the present disclosure;

[0056] Figure 2 A flowchart of a battery SOC estimation method provided by one embodiment of the present disclosure;

[0057] Figure 3 A flowchart of a battery SOC estimation method provided in yet another embodiment of the present disclosure;

[0058] Figure 4 A schematic structural diagram of a battery SOC estimation device provided in yet another embodiment of the present disclosure;

[0059] Figure 5 A schematic structural diagram of a control device provided in yet another embodiment of the present disclosure.

[0060] The above drawings illustrate specific embodiments of the present application, which will be described in more detail below. These drawings and the textual description are not intended to limit the scope of the present application in any way, but rather to illustrate the concepts of the present application to those skilled in the art by reference to specific embodiments. DETAILED DESCRIPTION

[0061] Exemplary embodiments will be described in detail herein, with examples illustrated in the accompanying drawings. In the following description, when referring to the drawings, identical numerals in different figures represent identical or similar elements, unless otherwise indicated. The embodiments described in the following exemplary embodiments are not intended to represent all embodiments consistent with the present application. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present application, as detailed in the appended claims.

[0062] The vehicle battery's SOC, or state of charge, is typically expressed as a percentage. For example, when the vehicle battery is fully charged, its theoretical SOC is 100%, and when the vehicle battery is fully discharged, its theoretical SOC is 0%. However, in practice, the SOC value does not reach these theoretical values, but rather varies between them. Accurate SOC determination is crucial to vehicle battery management. Inaccurate SOC determination can lead to discrepancies between the actual SOC and the estimated SOC, impacting battery charge and discharge management, vehicle operation, and other aspects, potentially posing a safety hazard.

[0063] In the prior art, the SOC of a vehicle battery is usually determined by estimation. The commonly used estimation method is to obtain the remaining capacity of the battery based on the difference between the real-time actual capacity and the actual capacity of the battery, and use the ratio of the remaining capacity to the total capacity as the estimated value of the battery SOC. However, this algorithm does not take into account issues such as ambient temperature and battery aging, and relies on high-precision sensors, resulting in high equipment costs, and there will always be a difference between the estimated value and the actual value of the battery SOC. There are also methods of calculation through models, such as the equivalent circuit method and the electrochemical model method, to improve the accuracy of the result estimation. However, these algorithms rely on high-precision battery models, have a large amount of calculation, and it is difficult for non-battery manufacturers to obtain the model parameters, which is not practical. In addition, there are SOC estimation algorithms based on neural networks to meet the needs of complex working conditions and have high accuracy, but these algorithms require a large amount of data and have too high hardware requirements, and they also have the problem of insufficient practicality.

[0064] The battery SOC estimation method, device, equipment and medium provided in this application fully consider the sampling error of the sensor and the result error caused by the ambient temperature during the battery SOC estimation process, and then correct the accumulated error in the calculation process through the Kalman filter algorithm, thereby maximizing the accuracy of the battery SOC estimation value.

[0065] Figure 1 This is a schematic diagram of an application scenario of the battery SOC estimation method provided in this application, such as Figure 1 As shown, in a vehicle, the onboard processor 100 obtains the charge and discharge current of the battery 120 through the battery sensor 110, and combines the data of the open circuit voltage and the ambient temperature to estimate the battery SOC.

[0066] It should be noted that Figure 1 The illustrated scenario includes only one or a specific number of vehicle-mounted processors, battery sensors, and batteries for illustration, but the present disclosure is not limited thereto. That is, the number of processors, battery sensors, and batteries can be arbitrary.

[0067] The following specific embodiments describe in detail the technical solution of the present application and how the technical solution of the present application solves the above-mentioned technical problems. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments. The embodiments of the present application will be described below in conjunction with the accompanying drawings.

[0068] Figure 2 This is a flow chart of a battery SOC estimation method provided by one embodiment of the present disclosure. Figure 2 As shown, the battery SOC estimation method provided by this embodiment includes the following steps:

[0069] S201 : Determine a real-time current corresponding to a target battery based on a sampling error of a battery sensor configured for the target battery and a sampled current acquired by the battery sensor.

[0070] Specifically, the executing entity of the embodiment of the present disclosure is an on-board processor, a battery management system, or a remote server connected to the on-board processor for calculating the vehicle battery SOC. For the convenience of description, they are collectively referred to as processors below.

[0071] When the vehicle is powered on, or the vehicle battery is powered on, the vehicle battery will undergo a continuous charging or discharging process, which will cause the vehicle battery to charge or discharge. The vehicle will be equipped with a battery sensor to monitor the current charging or discharging into the battery, and calculate the change in the battery's SOC value by integration based on the current size and the initial current value.

[0072] In this algorithm, the precision and accuracy of obtaining the numerical value of the battery's corresponding current is crucial to determining the SOC value.

[0073] In practice, the accuracy of current sensors is usually limited, and the current is usually collected by a Hall current sensor and needs to be transmitted to the processor through a bus. Therefore, corresponding collection noise exists in the process of data collection and transmission.

[0074] To ensure the accuracy of subsequent calculations, the acquisition noise must first be included in the calculations to avoid directly taking the data obtained by the processor as real data, which would affect the accuracy of subsequent SOC value determination.

[0075] The data error caused by acquisition noise, i.e., sampling error, can be reduced by taking the sampling error into account when determining the real-time acquired current (i.e., real-time current) (e.g., adding or subtracting the sampling error from the sampled current as the real-time current). This can significantly reduce the error in the subsequent SOC estimation calculation and improve the calculation accuracy.

[0076] S202 , input the real-time current into the ampere-hour integration algorithm, and output an estimated SOC value of the target battery.

[0077] Specifically, by inputting the real-time current into the ampere-hour integration algorithm, a preliminary calculated SOC estimate can be obtained.

[0078] However, there is still a deviation between this SOC estimate and the actual SOC value. This deviation is caused by factors such as ambient temperature and cumulative error. To eliminate this error, the SOC estimate needs to be further processed through the Kalman filter algorithm.

[0079] S203 , inputting the predetermined conversion relationship between SOC, open circuit voltage, and ambient temperature and the estimated SOC value into a Kalman filter algorithm, and outputting an estimated value of the battery SOC.

[0080] Specifically, in the Kalman filter algorithm, the influence of ambient temperature needs to be taken into consideration. Since the degree of influence of ambient temperature on the SOC of different batteries varies, it is also necessary to predetermine the relationship between the SOC of the target battery and the ambient temperature. Since the actual calculation also involves the open circuit voltage of the battery (the SOC converted by the open circuit voltage is compared with the subsequently calculated SOC value to correct the cumulative error), the relationship that needs to be determined is the relationship between SOC, open circuit voltage and ambient temperature, that is, the open circuit voltage conversion relationship based on temperature compensation.

[0081] The SOC estimate is obtained by combining the conversion relationship, real-time temperature and the aforementioned steps (in fact, in the ampere-hour integration algorithm, a more accurate initial SOC can also be calculated based on the conversion relationship and open-circuit voltage to improve the accuracy of the ampere-hour integration algorithm). When input into the Kalman filter algorithm, the SOC estimate can be corrected to obtain an SOC estimate that is more in line with reality.

[0082] Therefore, when determining the battery SOC estimation value, the influence of sampling errors caused by equipment reasons, ambient temperature and cumulative errors caused by integral operations can be significantly reduced, and the accuracy of the calculation results can be significantly improved. In addition, the calculation process does not require a complex neural network model or complex hardware to ensure sampling accuracy, which significantly improves the practicality of this solution.

[0083] The battery SOC estimation method provided by the disclosed embodiment determines the real-time current corresponding to the target battery by using the sampling error of the battery sensor configured based on the target battery and the sampled current obtained by the battery sensor. The real-time current is then input into an ampere-hour integration algorithm to output an estimated SOC value for the target battery. The predetermined SOC and the temperature-compensated open-circuit voltage conversion relationship and the estimated SOC value are then input into an adaptive extended Kalman filter algorithm to output an estimated battery SOC value. Thus, during the battery SOC estimation process, the sampling error of the sensor and the resulting error caused by the ambient temperature are fully considered. The accumulated error in the calculation process is then corrected by the Kalman filter algorithm, maximizing the accuracy of the battery SOC estimation value.

[0084] Figure 3 Flowchart of a battery SOC estimation method provided by another embodiment of the present disclosure. Figure 3 As shown, in Figure 2 Based on the embodiment shown, this embodiment further details the calculation process of the battery SOC estimation value, which includes the following steps:

[0085] S301: Obtain a preset sampling error parameter corresponding to a battery sensor.

[0086] The preset sampling error parameters include sampling error variance.

[0087] Specifically, when acquiring the real-time current, it is necessary to first acquire a preset sampling error parameter corresponding to the battery sensor. This parameter is a preset value corresponding to the battery sensor and can be acquired from the configuration parameters corresponding to the battery sensor.

[0088] The specific preset sampling error parameter used for calculation in this embodiment is the sampling error method.

[0089] Furthermore, in actual calculations, in addition to the sampling error variance, it is also necessary to obtain battery characteristic parameters (such as battery cell type, battery health status, etc.), and obtain real-time current data under battery charging and discharging conditions through battery sensors, vehicle-side open-circuit voltage data (such as open-circuit voltage meter), temperature and capacity characteristic data, etc., to ensure the accuracy of the subsequent SOC estimation value.

[0090] S302 , taking the difference between the sampled current and the sampling error variance as a correction value of the sampled current, and combining the correction value with the sampled current to obtain a real-time current.

[0091] Specifically, when the battery is discharging or charging, the sampling current output by the battery sensor can be added to the preset sampling error variance to correct the sampling error. However, since the actual fluctuation of the sampling current is not linear, the real-time current can be calculated by the following formula:

[0092] ,

[0093] Among them, I obs Used to indicate real-time current, I real It is used to represent the sampling current, and μ is used to represent the sampling error variance.

[0094] The above formula reflects the impact of sampling error on the sampling current during different battery discharge and charging processes, as well as during current changes. (Because as the sampling current changes, the right side of the above formula will also change to better reflect the impact of sampling error.)

[0095] S303 : Input the real-time current into the ampere-hour integration algorithm and output an estimated SOC value of the target battery.

[0096] Specifically, when performing specific ampere-hour integration, it is necessary to obtain the vehicle's real-time current under a specific operating condition and then substitute it into the ampere-hour integration algorithm for calculation. For example, the real-time current when the vehicle is continuously running and discharging, or the real-time current when the vehicle is parked and charging.

[0097] In the embodiment of the present disclosure, the ampere-hour integration algorithm is:

[0098] ,

[0099] Among them, SOC(t) represents the estimated SOC at time t, I(τ) is used to represent the real-time current at time τ, SOC0 is used to represent the initial SOC of the target battery, and the initial SOC is determined by the open circuit voltage conversion relationship based on temperature compensation, C normal It is used to indicate the rated capacity, and t represents the duration that the target battery is in the charging or discharging state.

[0100] Specifically, SOC0 is actually the starting SOC value collected under the same operating condition, or it can be the final SOC estimate calculated under the previous operating condition. By limiting the calculation to the same operating condition, the accuracy of subsequent SOC estimate calculations can be guaranteed.

[0101] In some embodiments, when the vehicle operating condition changes, the SOC value finally obtained in the previous state needs to be used as the new SOC0, and the time t needs to be reset to zero to recalculate the SOC estimation value under the operating condition to ensure the accuracy of the calculation.

[0102] To ensure the accuracy of the ampere-hour integration algorithm, the corresponding SOC0 is calculated through the open-circuit voltage conversion formula based on temperature compensation, combined with the real-time detected open-circuit voltage and real-time ambient temperature. Therefore, on the basis of the traditional current ampere-hour integration, the open-circuit voltage conversion formula based on temperature compensation is corrected to improve the calculation accuracy of the open-circuit voltage, thereby ensuring the calculation accuracy of the SOC estimation value.

[0103] S304 : Determine the actual capacity of the target battery based on the preset temperature characteristics, cell type, battery health status, and rated capacity of the target battery.

[0104] Specifically, after obtaining a preliminary SOC estimate in the above steps, the SOC estimate needs to be revised. First, the current actual capacity of the vehicle needs to be further clarified.

[0105] Furthermore, the actual capacity is expressed as:

[0106] ,

[0107] Among them, CAP represents the actual capacity of the target battery, F represents the preset temperature characteristic value, temp represents the real-time ambient temperature, celltype represents the type of battery cell, C normal It is used to indicate the rated capacity, and SOH indicates the battery state of health.

[0108] By determining the real-time ambient temperature under the current operating conditions, combining the battery cell type, and querying the preset temperature characteristic value stored in the processor, the corresponding F value can be determined. Combined with the currently determined battery health status value (this value can be calculated by the processor, and there are mature algorithms for calculating battery health status in related technologies. The calculation method is not included in this solution, so it is not described in detail here. Those skilled in the art can choose any algorithm to determine SOH according to actual conditions, and there is no restriction here) and the rated capacity, the actual battery capacity under the current conditions can be obtained.

[0109] In subsequent calculations, CAP is used instead of C normal Calculation is performed, although the formula still uses C when battery capacity is involved normal However, in actual calculations, CAP is used to ensure that the calculation results are consistent with the actual situation.

[0110] In step S305 , the actual capacity, the temperature-compensated open circuit voltage conversion equation, and the estimated SOC value are input into an adaptive extended Kalman filter algorithm to output the battery SOC estimate.

[0111] Specifically, when the above results are obtained, the SOC estimation value can be corrected by using the adaptive extended Kalman filter algorithm.

[0112] In practical applications, the processor usually predetermines the open circuit voltage conversion relationship based on temperature compensation rather than determining it in real time. The specific determination process is described here, which includes the following steps:

[0113] Step A1: Obtain the open circuit voltage corresponding to different ambient temperatures and different SOC values, and use the open circuit voltage, the corresponding ambient temperature, and the corresponding SOC value as sample data.

[0114] Specifically, for the same battery, the changing relationship between its SOC value and open circuit voltage at different ambient temperatures can be obtained, thereby obtaining an array containing open circuit voltage, ambient temperature, and SOC value. These arrays can then be used as sample data to determine the conversion relationship.

[0115] Step A2: Divide the sample data into at least three interval groups based on the distribution range of the SOC value.

[0116] Specifically, since the relationship between open circuit voltage, SOC value and ambient temperature varies under different SOC values ​​(for example, the open circuit voltage of some batteries is more affected by temperature when the SOC value is low), it is necessary to group the sample data according to the SOC value distribution range.

[0117] For example, the sample data corresponding to the SOC value of 0-10% can be divided into the first group, the sample data corresponding to 10%-90% can be divided into the second group, and the sample data corresponding to 90%-100% can be divided into the third group.

[0118] In practical applications, further subdivision can be done (such as further splitting the third group into 90%~95% and 95%~100%), or the range of each group can be adjusted (such as adjusting to 85%~100%) to suit the specific characteristics of the battery.

[0119] Step A3: Substitute the sample data of each interval group into the fitting formula to determine the fitting formula parameters corresponding to each interval group.

[0120] The fitting formula is used to express the calculation relationship between the open circuit voltage and the corresponding SOC value and ambient temperature.

[0121] Specifically, the sample data of each group can be fitted separately to obtain the fitting formula parameters corresponding to each segment, and then obtain the fitting formula corresponding to the group.

[0122] Since this process is not completed in real time, it can actually be processed through a special data fitting program to ensure the accuracy of the fitting.

[0123] Furthermore, the fitting formula is:

[0124] ,

[0125] Wherein, OCV is used to represent the open circuit voltage, T is used to represent the ambient temperature, and a0, a1, a2, a3, b1, and b2 are fitting formula parameters.

[0126] Within a certain SOC range, the open circuit voltage, ambient temperature and SOC value usually have a relatively fixed correlation. Therefore, the corresponding fitting formula can be better obtained through the above formula.

[0127] Step A4: Combine the fitting formulas corresponding to each interval group to obtain a conversion relationship.

[0128] The conversion formula includes at least three fitting formulas corresponding to different SOC distribution intervals.

[0129] Specifically, the fitting formulas for each group are combined to produce segmented fitting formulas corresponding to each SOC range, known as multi-segment SOC-OCV polynomials (OCV is open-circuit voltage). This provides a temperature-compensated open-circuit voltage conversion equation that accurately reflects the relationship between SOC, open-circuit voltage, and ambient temperature. To more intuitively demonstrate the relationship between SOC and open-circuit voltage, the conversion equation can be expressed as a multi-segment OCV (open-circuit voltage) curve, reflecting the corresponding open-circuit voltage distribution at different temperatures and SOC levels.

[0130] In practical applications, each SOC-OCV polynomial can be stored as a model separately, and the corresponding model can be selected for separate calculation based on the range of the SOC value determined in real time, and there is no weight difference between the models.

[0131] Therefore, we can combine the voltage change characteristics of SOC at different stages, determine the range of SOC, and select the algorithm (i.e., fitting formula) of the corresponding range to execute, so as to adapt to the needs of different working conditions.

[0132] In the adaptive extended Kalman filter algorithm, the conversion relationship can be used to calculate the Kalman gain and to determine SOC0. By determining the open-circuit voltage and the real-time ambient temperature in real time, the corresponding value can be calculated. The calculation result can also be used to determine SOC0 in the ampere-hour integration algorithm.

[0133] On this basis, the adaptive extended Kalman filter algorithm can be used for calculation. Refer to the following steps to explain this part:

[0134] Step B1: Use SOC0 as the initial value and determine the corresponding initial parameters.

[0135] Among them, the initial parameters include the covariance matrix P0, process noise Q, and observation noise R. The initial parameters are all unit matrices, and the process noise and observation noise are both covariance matrices.

[0136] Specifically, for the Kalman filter calculation corresponding to each working condition, the SOC0 selected when using the ampere-hour integration algorithm can be used as the initial value, so as to start from the initial value and perform the filter correction.

[0137] The initial parameters are used to correct and calculate the initial values ​​and subsequent predicted values, and are updated synchronously with the calculations. Initially, P0 is set to the minimum unit matrix, and all other matrices are also unit matrices.

[0138] Step B2: Determine the SOC prediction value and prediction parameters after the set time based on the initial value and initial parameters.

[0139] Among them, the prediction parameters include the prediction covariance matrix.

[0140] Specifically, the SOC prediction value is:

[0141] ,

[0142] Among them, SOC pre (k) represents the SOC prediction value corresponding to the k-th step, SOC(k-1) represents the SOC prediction value of the previous step corresponding to SOCpre(k), η represents the charge and discharge efficiency, I represents the real-time current, and Δt represents the time between adjacent steps.

[0143] After each step (i.e., the set unit time length, or the sampling time interval of the sampling current), the specific SOC prediction value and prediction parameters will be calculated.

[0144] For ease of calculation, the discharge efficiency of η is set to 1 by default, and the charge efficiency is set to 0.95 by default. These can be adjusted in actual situations.

[0145] The prediction parameters are:

[0146] ,

[0147] Among them, P pre (k) represents the prediction covariance matrix corresponding to the k-th step, P(k-1) represents P pre (k) represents the prediction covariance matrix of the previous step corresponding to Q(k), and Q(k) represents the process noise covariance matrix corresponding to the k-th step.

[0148] Each time the SOC prediction value is calculated, the parameters of the covariance matrix also need to be updated.

[0149] Step B3: Determine the dynamic parameters corresponding to the SOC prediction value based on the actual capacity and the conversion formula.

[0150] The dynamic parameters include the Jacobian matrix and the Kalman gain.

[0151] Specifically, the Jacobian matrix is:

[0152] ,

[0153] Among them, H k represents the Jacobian matrix corresponding to the k-th step size, and f represents the conversion relationship;

[0154] Specifically, the conversion relationship obtained in the above steps can be used ocv =f(SOC,T). Under certain working conditions and current SOC prediction values, the corresponding conversion formula can be selected. By taking the partial derivative of the SOC in the conversion formula, the Jacobian matrix can be obtained.

[0155] The Kalman gain is:

[0156] ,

[0157] Among them, K k represents the Kalman gain corresponding to the k-th step length, and R(k) represents the observation noise covariance matrix corresponding to the k-th step length.

[0158] Specifically, the corresponding Kalman gain can be calculated by combining the Jacobian matrix, the prediction covariance matrix, and the observation noise covariance matrix.

[0159] Step B4: Update the SOC prediction value based on the dynamic parameters.

[0160] Specifically, the SOC prediction value is updated as follows:

[0161] ,

[0162] Among them, U up (k) represents the updated open circuit voltage corresponding to the kth step, U pre (k) represents the predicted open circuit voltage corresponding to the kth step, U ocv (k) represents the open circuit voltage determined based on the conversion relationship corresponding to the k-th step, SOC(k) represents the updated SOC prediction value corresponding to the k-th step, SOC pre (k) represents the SOC prediction value corresponding to the k-th step.

[0163] Specifically, the open circuit voltage and SOC prediction values ​​can be corrected through the Kalman gain, and updated open circuit voltage and SOC prediction values ​​can be calculated.

[0164] Step B5: adaptively update the initial parameters based on the difference between the updated SOC prediction value and the SOC estimated value.

[0165] Specifically, in combination with the correction of the SOC prediction value, each initial parameter needs to be dynamically and adaptively updated. The algorithm can be expressed as follows:

[0166] ,

[0167] Among them, b is the forgetting factor, d is the weight parameter, k is the step size, Q(k-1) represents the process noise covariance matrix corresponding to the k-1th step size, K k ′ represents the derivative of the Kalman gain corresponding to the k-th step, epsilon represents the observation residual, and epsilon′ represents the derivative of the observation residual. The observation residual is used to represent the difference between the SOC estimate value obtained based on the open-circuit voltage in the k-th step and the SOC estimation value.

[0168] Specifically, by combining the observation residual and Kalman gain to correct the initial parameters, the weight parameters and noise variance (including process noise and observation noise) can be adaptively updated in each calculation according to the working conditions, thereby achieving noise adaptation under different working conditions and thus improving the accuracy of error correction in the calculation process under different working conditions.

[0169] Step B6: Use the updated SOC prediction value as the SOC estimation value.

[0170] Specifically, through the above processing, the calculation of the SOC value corresponding to the open circuit voltage can be combined with the error correction of noise and temperature, without introducing an equivalent circuit model or a complex neural network model, thereby reducing the amount of calculation.

[0171] The battery SOC estimation method provided in the embodiment of the present disclosure corrects the measurement error of the real-time current by combining the sampling error variance, and then combines the open-circuit voltage conversion relationship under different temperatures and different SOCs to ensure the calculation accuracy of the open-circuit voltage. Then, through the combination of the adaptive noise Kalman filter algorithm and the ampere-hour integration algorithm, the process error is dynamically adjusted and the error correction is performed. The battery health is also combined to maximize the accuracy of the calculated SOC estimate while reducing the amount of calculation.

[0172] Figure 4 This is a schematic diagram of the structure of a battery SOC estimation device provided by an embodiment of the present disclosure. Figure 4 As shown, the battery SOC estimation device 400 includes:

[0173] A determination module 410 is configured to determine a real-time current corresponding to a target battery based on a sampling error of a battery sensor configured for the target battery and a sampled current obtained by the battery sensor;

[0174] An estimation module 420 is configured to input the real-time current into an ampere-hour integration algorithm and output an estimated SOC value of the target battery;

[0175] Output module 430 is used to input a predetermined SOC and a temperature-compensated open-circuit voltage conversion relationship and an estimated SOC value into an adaptive extended Kalman filter algorithm, and output an estimated value of the battery SOC, wherein the adaptive extended Kalman filter algorithm includes adaptively updated process noise and observation noise.

[0176] Optionally, the determination module 410 is specifically used to obtain preset sampling error parameters corresponding to the battery sensor, where the preset sampling error parameters include sampling error variance; use the difference between the sampling current and the sampling error variance as a correction value of the sampling current, and combine the correction value with the sampling current to obtain the real-time current.

[0177] Optionally, the determination module 410 specifically includes: the real-time current I obs for:

[0178] ,

[0179] Among them, I real It is used to represent the sampling current, and μ is used to represent the sampling error variance.

[0180] Optionally, the estimation module 420 specifically includes an ampere-hour integration algorithm:

[0181] ,

[0182] Among them, SOC(k-1) represents the estimated SOC at time t, I(τ) is used to represent the real-time current at time τ, SOC0 is used to represent the initial SOC of the target battery, and the initial SOC is determined by the open circuit voltage conversion relationship based on temperature compensation, C normal It is used to indicate the rated capacity, and t represents the duration that the target battery is in the charging or discharging state.

[0183] Optionally, the output module 430 is specifically used to determine an open circuit voltage conversion equation based on temperature compensation in the following manner: obtaining the open circuit voltage corresponding to different ambient temperatures and different SOC values, and using the open circuit voltage, the corresponding ambient temperature and the corresponding SOC value as sample data; dividing the sample data into at least three interval groups based on the distribution range of the SOC value; substituting the sample data of each interval group into the fitting formula respectively, and determining the fitting formula parameters corresponding to each interval group, wherein the fitting formula is used to express the calculation relationship between the open circuit voltage and the corresponding SOC value and the ambient temperature; combining the fitting formulas corresponding to each interval group to obtain a conversion equation, wherein the conversion equation includes at least three fitting formulas corresponding to different SOC distribution intervals.

[0184] Optionally, the output module 430 specifically includes a fitting formula:

[0185] ,

[0186] Wherein, OCV is used to represent the open circuit voltage, T is used to represent the ambient temperature, and a0, a1, a2, a3, b1, and b2 are fitting formula parameters.

[0187] Optionally, the output module 430 is specifically used to determine the actual capacity of the target battery based on the preset temperature characteristics, cell type, battery health status and rated capacity of the target battery; input the actual capacity, the open circuit voltage conversion relationship based on temperature compensation, and the SOC estimated value into the adaptive extended Kalman filter algorithm to output the battery SOC estimation value.

[0188] Optionally, the output module 430 specifically includes: the actual capacity is expressed as:

[0189] ,

[0190] Among them, CAP represents the actual capacity of the target battery, F represents the preset temperature characteristic value, temp represents the real-time ambient temperature, celltype represents the type of battery cell, C normal It is used to indicate the rated capacity, and SOH indicates the battery state of health.

[0191] Optionally, the output module 430 is specifically used to take SOC0 as the initial value and determine the corresponding initial parameters, wherein the initial parameters include the covariance matrix P0, the process noise Q, and the observation noise R, the initial parameters are all unit matrices, and the process noise and the observation noise are both covariance matrices; based on the initial value and the initial parameters, determine the SOC prediction value and prediction parameters after the set time, the prediction parameters include the prediction covariance matrix; based on the actual capacity and the conversion relationship, determine the dynamic parameters corresponding to the SOC prediction value, the dynamic parameters include the Jacobian matrix and the Kalman gain; based on the dynamic parameters, update the SOC prediction value; based on the difference between the updated SOC prediction value and the SOC estimated value, adaptively update the initial parameters; and use the updated SOC prediction value as the SOC estimated value.

[0192] Optionally, the output module 430 specifically includes: the SOC prediction value is:

[0193] ,

[0194] Among them, SOC pre (k) represents the SOC prediction value corresponding to the k-th step, SOC(k-1) represents the SOC prediction value of the previous step corresponding to SOCpre(k), η represents the charge and discharge efficiency, I represents the real-time current, and Δt represents the time between adjacent steps;

[0195] The prediction parameters are:

[0196] ,

[0197] Among them, P pre (k) represents the prediction covariance matrix corresponding to the k-th step, P(k-1) represents P pre (k) is the prediction covariance matrix of the previous step corresponding to the k-th step, Q(k) represents the process noise covariance matrix corresponding to the k-th step;

[0198] The Jacobian matrix is:

[0199] ,

[0200] Among them, H k represents the Jacobian matrix corresponding to the k-th step size, and f represents the conversion relationship;

[0201] The Kalman gain is:

[0202] ,

[0203] Among them, K k represents the Kalman gain corresponding to the k-th step length, and R(k) represents the observation noise covariance matrix corresponding to the k-th step length;

[0204] The SOC prediction value is updated as:

[0205] ,

[0206] Among them, U up (k) represents the updated open circuit voltage corresponding to the kth step, U pre (k) represents the predicted open circuit voltage corresponding to the kth step, U ocv (k) represents the open circuit voltage determined based on the conversion relationship corresponding to the k-th step, SOC(k) represents the updated SOC prediction value corresponding to the k-th step, SOC pre (k) represents the SOC prediction value corresponding to the k-th step length;

[0207] The adaptive update of the initial parameters is expressed as:

[0208] ,

[0209] Among them, b is the forgetting factor, d is the weight parameter, k is the step size, Q(k-1) represents the process noise covariance matrix corresponding to the k-1th step size, K k ′ represents the derivative of the Kalman gain corresponding to the k-th step, epsilon represents the observation residual, and epsilon′ represents the derivative of the observation residual. The observation residual is used to represent the difference between the SOC estimate value obtained based on the open-circuit voltage in the k-th step and the SOC estimation value.

[0210] The functions and principles of each module in this embodiment can be found in the aforementioned method embodiment and will not be repeated here.

[0211] Figure 5 A schematic diagram of the structure of a control device provided in one embodiment of the present disclosure is shown in FIG. Figure 5 As shown, the control device 500 includes: a memory 510 and a processor 520.

[0212] The memory 510 stores a computer program that can be executed by at least one processor 520. The computer program is executed by at least one processor 520 to enable the control device to implement the battery SOC estimation method provided in any of the above embodiments.

[0213] The memory 510 and the processor 520 may be connected via a bus 530 .

[0214] The relevant instructions can be understood by referring to the relevant descriptions and effects corresponding to the method embodiments, which will not be repeated here.

[0215] An embodiment of the present disclosure provides a computer-readable storage medium having a computer program stored thereon. The computer program is executed by a processor to implement the battery SOC estimation method provided by any of the above method embodiments.

[0216] The computer-readable storage medium may be a ROM, a random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, or the like.

[0217] An embodiment of the present disclosure provides a computer program product, which includes computer-executable instructions. When the computer-executable instructions are executed by a processor, they are used to implement the battery SOC estimation method provided in any of the above embodiments.

[0218] In the above embodiments, the description of each embodiment has its own emphasis. For parts not described in detail in a particular embodiment, please refer to the relevant description of other embodiments. The technical features of the above embodiments can be combined in any way. To keep the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0219] Those skilled in the art will readily appreciate other embodiments of the present application after considering the specification and practicing the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of the present application that follow the general principles of the present application and include common knowledge or customary techniques in the art not disclosed herein. The description and examples are to be considered as exemplary only, and the true scope and spirit of the present application are indicated by the following claims.

[0220] It should be understood that the present application is not limited to the exact structure described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present application is limited only by the appended claims.

Claims

1. A battery SOC estimation method, characterized in that: include: Determining a real-time current corresponding to the target battery based on a sampling error of a battery sensor configured for the target battery and a sampled current obtained by the battery sensor; Inputting the real-time current into an ampere-hour integration algorithm and outputting an estimated SOC value of the target battery; Inputting a predetermined SOC and temperature-compensated open circuit voltage conversion relationship and the SOC estimate into an adaptive extended Kalman filter algorithm to output a corrected battery SOC estimate, wherein the adaptive extended Kalman filter algorithm includes adaptively updated process noise and observation noise; Determining a real-time current corresponding to the target battery based on a sampling error of a battery sensor configured for the target battery and a sampled current obtained by the battery sensor includes: Obtaining a preset sampling error parameter corresponding to the battery sensor, wherein the preset sampling error parameter includes a sampling error variance; Using the difference between the sampled current and the sampling error variance as a correction value of the sampled current, and combining the correction value with the sampled current to obtain the real-time current; The real-time current is calculated by the following formula: ， Among them, I obs Used to indicate real-time current, I real It is used to represent the sampling current, and μ is used to represent the sampling error variance.

2. The method according to claim 1, characterized in that The ampere-hour integration algorithm is: ， Wherein, SOC(t) represents the estimated SOC at time t, I(τ) represents the real-time current at time τ, SOC0 represents the initial SOC of the target battery, and the initial SOC is determined by the open circuit voltage conversion relationship based on temperature compensation, C normal It is used to indicate the rated capacity, and t represents the duration that the target battery is in the charging or discharging state.

3. The method according to claim 2, characterized in that The open circuit voltage conversion relationship based on temperature compensation is determined as follows: Obtaining open circuit voltages corresponding to different SOC values ​​at different ambient temperatures, and using the open circuit voltages, corresponding ambient temperatures, and corresponding SOC values ​​as sample data; Dividing the sample data into at least three interval groups based on the distribution range of the SOC value; Substituting the sample data of each interval group into the fitting formula respectively to determine the fitting formula parameters corresponding to each interval group, wherein the fitting formula is used to express the calculation relationship between the open circuit voltage and the corresponding SOC value and the ambient temperature; The fitting formulas corresponding to each interval grouping are combined to obtain the conversion relationship formula, wherein the conversion relationship formula includes at least three fitting formulas corresponding to different SOC distribution intervals.

4. The method according to any one of claims 1 to 3, characterized in that Inputting a predetermined SOC and a temperature-compensated open circuit voltage conversion relationship and the battery SOC estimation value into an adaptive extended Kalman filter algorithm to output the estimated value of the battery SOC includes: Determining an actual capacity of the target battery based on preset temperature characteristics, cell type, battery health status, and rated capacity of the target battery; The actual capacity, the temperature-compensated open-circuit voltage conversion equation, and the SOC estimation value are input into the adaptive extended Kalman filter algorithm, and the battery SOC estimation value is output.

5. The method according to claim 4, characterized in that Inputting the actual capacity, the temperature-compensated open circuit voltage conversion equation, and the estimated SOC value into the adaptive extended Kalman filter algorithm to output the battery SOC estimate includes: Taking SOC0 as the initial value and determining the corresponding initial parameters, wherein the initial parameters include the covariance matrix P0, the process noise Q, and the observation noise R, the initial parameters are all unit matrices, and the process noise and the observation noise are both covariance matrices; Determining a predicted SOC value and predicted parameters after a set time period based on the initial value and initial parameters, wherein the predicted parameters include a predicted covariance matrix; Determining dynamic parameters corresponding to the SOC prediction value based on the actual capacity and the conversion relationship, the dynamic parameters including a Jacobian matrix and a Kalman gain; updating the SOC prediction value based on the dynamic parameter; Adaptively updating the initial parameters based on a difference between the updated SOC predicted value and the SOC estimated value; The updated SOC prediction value is used as the SOC estimation value.

6. The method according to claim 5, characterized in that The adaptive update of the initial parameters is expressed as: ， Among them, b is the forgetting factor, d is the weight parameter, k is the step size, Q(k-1) represents the process noise corresponding to the k-1th step size, K k ′ represents the derivative of the Kalman gain corresponding to the kth step, epsilon represents the observation residual, and epsilon′ represents the derivative of the observation residual. The observation residual is used to represent the difference between the SOC estimated value obtained based on the open circuit voltage in the kth step and the SOC estimated value.

7. A battery SOC estimation device, characterized in that: The battery SOC estimation device includes: a determination module, configured to determine a real-time current corresponding to the target battery based on a sampling error of a battery sensor configured for the target battery and a sampled current obtained by the battery sensor; an estimation module, configured to input the real-time current into an ampere-hour integration algorithm and output an estimated SOC value of the target battery; an output module, configured to input a predetermined SOC and a temperature-compensated open circuit voltage conversion relationship, and the estimated SOC value, into an adaptive extended Kalman filter algorithm, and output an estimated value of the battery SOC, wherein the adaptive extended Kalman filter algorithm includes adaptively updated process noise and observation noise; The determining module is specifically configured to: Obtaining a preset sampling error parameter corresponding to the battery sensor, wherein the preset sampling error parameter includes a sampling error variance; Using the difference between the sampled current and the sampling error variance as a correction value of the sampled current, and combining the correction value with the sampled current to obtain the real-time current; The real-time current is calculated by the following formula: ， Among them, I obs Used to indicate real-time current, I real It is used to represent the sampling current, and μ is used to represent the sampling error variance.

8. A control device, characterized in that: include: at least one processor; and a memory communicatively coupled to the at least one processor; The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the control device to perform the battery SOC estimation method according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer-executable instructions, which are used to implement the battery SOC estimation method according to any one of claims 1 to 6 when executed by a processor.

Images