Battery state prediction method and device, storage medium, battery and vehicle

Abstract

The invention relates to a battery state prediction method and device, a storage medium, a battery and a vehicle, and the method comprises the steps: carrying out the updating of different battery state parameters based on different updating frequencies, and predicting the state of the battery based on the updated battery state parameters, the battery state parameters comprise at least two of charge state parameters, battery model parameters and capacity parameters. According to the method and the device, updating requirements of different battery state parameters can be better adapted, so that the updated battery state parameters can be more accurately used for updating the battery state parameters.

Description

Technical Field

[0001] This application relates to the field of battery state monitoring, and more particularly to a battery state prediction method, device, storage medium, battery, and vehicle. Background Technology

[0002] With the development of electric and hybrid vehicles, the accuracy of monitoring the state of on-board batteries affects vehicle performance. In actual battery use, as the battery's State of Health (SOH) changes, various state parameters used to calculate the battery's state, such as state of charge, capacity, and battery model parameters, will also change, thus affecting the calculation of the actual battery state. Therefore, it is often necessary to correct these state parameters.

[0003] However, most current solutions rely on aging curves to correct and adjust various state parameters during battery use. The correction effect is not ideal, and the parameters cannot be corrected in real time. Summary of the Invention

[0004] This application provides a battery state prediction method, device, storage medium, battery, and vehicle, aiming to solve the problem of inaccurate battery state prediction in related technologies.

[0005] In a first aspect, this application provides a battery state prediction method, including:

[0006] Different battery state parameters are updated at different update frequencies to predict battery state based on the updated battery state parameters. The battery state parameters include at least two of the following: state of charge parameters, battery model parameters, and capacity parameters.

[0007] As a further feasible implementation of this application, the step of updating different battery state parameters based on different update frequencies includes:

[0008] Different battery state parameters are updated when the cumulative duration and / or change in charge during the battery charging and discharging process meet different conditions.

[0009] As a further feasible implementation of this application, the updating of different battery state parameters when the cumulative duration and / or charge change during the battery charging and discharging process meet different conditions includes:

[0010] When the cumulative duration of battery charging and discharging reaches a first preset duration and / or the change in charge reaches a first preset change, the capacity parameter in the battery state parameters is updated; or

[0011] When the cumulative duration of battery charging and discharging reaches a second preset duration and / or the change in charge reaches a second preset change, the battery model parameters in the battery state parameters are updated; or

[0012] When the cumulative duration of the battery charging and discharging process reaches a third preset duration and / or the change in charge reaches a third preset change, the state of charge parameters in the battery state parameters are updated.

[0013] As a further feasible implementation of this application, the first preset duration is longer than the second preset duration, the second preset duration is longer than the third preset duration, the first preset change is greater than the second preset change, and the second preset change is greater than the third preset change.

[0014] As a further feasible implementation of this application, the step of updating different battery state parameters includes:

[0015] Based on the previous capacity parameters and the prior estimates of the state of charge parameters, the capacity parameters in the battery state parameters are updated; or

[0016] Based on the battery model parameters before the update and the prior estimates of the state of charge parameters, the battery model parameters in the battery state parameters are updated; or

[0017] Based on the previous state of charge parameters, the prior estimated parameters of the capacity parameters, and the prior estimated values ​​of the battery model parameters, the state of charge parameters in the battery state parameters are updated.

[0018] As a further feasible implementation of this application, updating the capacity parameter in the battery state parameters based on the previous capacity parameter and the prior estimate of the state of charge parameter includes:

[0019] The prior estimate of the capacity parameter is obtained by performing a preset filtering algorithm to perform a prior estimate of the capacity parameter before the update.

[0020] Based on the prior estimate of the state of charge parameter, the prior estimate of the capacity parameter is estimated posteriorly to obtain the updated value of the capacity parameter in the battery state parameters.

[0021] As a further feasible implementation of this application, the step of performing a priori estimation of the capacity parameter before the update using a preset filtering algorithm to obtain the priori estimate value of the capacity parameter includes:

[0022] Based on the covariance matrix of the capacity parameter before the update, multiple prior distribution values ​​of the capacity parameter are obtained.

[0023] Based on multiple prior distribution values ​​of the capacity parameter, a prior estimate of the capacity parameter and a prior estimate of the covariance matrix of the capacity parameter are obtained.

[0024] If the prior estimate of the covariance matrix of the capacity parameter and / or the prior estimate of the capacity parameter meet preset conditions, the prior estimate of the capacity parameter is updated.

[0025] As a further feasible implementation of this application, obtaining a prior estimate of the capacity parameter based on multiple prior distribution values ​​of the capacity parameter, and a prior estimate of the covariance matrix of the capacity parameter, includes:

[0026] Based on the mean weight corresponding to the prior distribution value of each capacity parameter, the results obtained by processing the prior distribution value of the capacity parameter through the capacity filtering state equation are weighted to obtain the prior estimate value of the capacity parameter.

[0027] Based on the covariance weights corresponding to the prior distribution values ​​of each capacity parameter, the differences between the prior distribution values ​​and the prior estimates of the capacity parameters are weighted to obtain the prior estimates of the covariance matrix of the capacity parameters.

[0028] As a further feasible implementation of this application, updating the prior estimate of the capacity parameter when the prior estimate of the covariance matrix of the capacity parameter and / or the prior estimate of the capacity parameter satisfies a preset condition includes:

[0029] When the prior estimate of the covariance matrix of the capacity parameter is a pairwise matrix, a positive definite matrix, and / or the prior estimate of the covariance matrix of the capacity parameter is within a preset covariance boundary range, and / or the prior estimate of the capacity parameter is within a preset capacity boundary range, the prior estimate of the capacity parameter is updated.

[0030] As a further feasible implementation of this application, the step of performing a posteriori estimation on the prior estimate of the capacity parameter based on the prior estimate of the state of charge parameter to obtain the updated value of the capacity parameter in the battery state parameters includes:

[0031] Based on the prior estimates of the capacity parameter and the prior estimates of the covariance matrix of the capacity parameter, multiple posterior distribution values ​​of the capacity parameter are obtained.

[0032] The first measurement mean and the first measurement covariance matrix are obtained by processing multiple posterior distribution values ​​of the capacity parameter and the prior estimate value of the state of charge parameter through the capacity filtering measurement equation.

[0033] Based on the first measurement mean and the first gain matrix corresponding to the first measurement covariance matrix, the prior estimate of the capacity parameter and the prior estimate of the covariance matrix of the capacity parameter are updated to obtain the updated value of the capacity parameter. The updated value of the capacity parameter includes the posterior estimate of the capacity parameter and the posterior estimate of the covariance matrix of the capacity parameter.

[0034] As a further feasible implementation of this application, updating the battery model parameters in the battery state parameters based on the battery model parameters before the update and the prior estimates of the state of charge parameters includes:

[0035] The battery model parameters before the update are estimated using a preset filtering algorithm to obtain the prior estimate values ​​of the battery model parameters.

[0036] Based on the prior estimates of the state of charge parameters, the prior estimates of the battery model parameters are posteriorly estimated to obtain the updated values ​​of the battery model parameters in the battery state parameters.

[0037] As a further feasible implementation of this application, the step of performing prior estimation on the battery model parameters before the update using a preset filtering algorithm to obtain prior estimated values ​​of the battery model parameters includes:

[0038] Based on the covariance matrix of the battery model parameters before the update, multiple prior distribution values ​​of the battery model parameters are obtained.

[0039] Based on multiple prior distribution values ​​of the battery model parameters, prior estimates of the battery model parameters and prior estimates of the covariance matrix of the battery model parameters are obtained.

[0040] If the prior estimate of the covariance matrix of the battery model parameters and / or the prior estimate of the battery model parameters meets a preset condition, the prior estimate of the battery model parameters is updated.

[0041] As a further feasible implementation of this application, obtaining prior estimates of the battery model parameters and prior estimates of the covariance matrix of the battery model parameters based on multiple prior distribution values ​​of the battery model parameters includes:

[0042] Based on the mean weight corresponding to the prior distribution value of each battery model parameter, the results obtained by processing the prior distribution values ​​of the battery model parameters through the battery model filtering state equation are weighted to obtain the prior estimate value of the battery model parameter.

[0043] Based on the covariance weights corresponding to the prior distribution values ​​of each battery model parameter, the differences between the prior distribution values ​​and the prior estimates of the battery model parameters are weighted to obtain the prior estimates of the covariance matrix of the battery model parameters.

[0044] As a further feasible implementation of this application, the step of updating the prior estimate of the battery model parameters when the prior estimate of the covariance matrix of the battery model parameters and / or the prior estimate of the battery model parameters meets a preset condition includes:

[0045] When the prior estimate of the covariance matrix of the battery model parameters is a pairwise matrix, a positive definite matrix, and / or the prior estimate of the covariance matrix of the battery model parameters is within a preset covariance boundary range, and / or the prior estimate of the battery model parameters is within a preset model parameter boundary range, the prior estimate of the battery model parameters is updated.

[0046] As a further feasible implementation of this application, the step of performing a posteriori estimation on the prior estimates of the battery model parameters based on the prior estimates of the state of charge parameters to obtain updated values ​​of the battery model parameters in the battery state parameters includes:

[0047] Based on the prior estimates of the battery model parameters and the prior estimates of the covariance matrix of the battery model parameters, multiple posterior distribution values ​​of the battery model parameters are obtained.

[0048] The battery model filtering measurement equation is used to process multiple posterior distribution values ​​of the battery model parameters and prior estimates of the state of charge parameters to obtain the second measurement mean and the second measurement covariance matrix.

[0049] Based on the second measurement mean and the second gain matrix corresponding to the second measurement covariance matrix, the prior estimates of the battery model parameters and the prior estimates of the covariance matrix of the battery model parameters are updated to obtain updated values ​​of the battery model parameters. The updated values ​​of the battery model parameters include the posterior estimates of the battery model parameters and the posterior estimates of the covariance matrix of the battery model parameters.

[0050] As a further feasible implementation of this application, updating the state-of-charge parameters in the battery state parameters based on the previous state-of-charge parameters, the prior estimated parameters of the capacity parameters, and the prior estimated values ​​of the battery model parameters includes:

[0051] The prior estimate of the charge state parameters before the update is obtained by using a preset filtering algorithm.

[0052] Based on the prior estimated parameters of the capacity parameter and the prior estimated values ​​of the battery model parameters, the prior estimated values ​​of the state of charge parameter are posteriorly estimated to obtain the updated values ​​of the state of charge parameter in the battery state parameters.

[0053] As a further feasible implementation of this application, the step of performing prior estimation of the charge state parameters before updating using a preset filtering algorithm to obtain prior estimated values ​​of the charge state parameters includes:

[0054] Based on the covariance matrix of the charge state parameters before the update, multiple prior distribution values ​​of the charge state parameters are obtained.

[0055] Based on multiple prior distribution values ​​of the charged state parameters, prior estimates of the charged state parameters and prior estimates of the covariance matrix of the charged state parameters are obtained.

[0056] If the prior estimate of the covariance matrix of the charge state parameter and / or the prior estimate of the charge state parameter meets a preset condition, the prior estimate of the charge state parameter is updated.

[0057] As a further feasible implementation of this application, obtaining the prior estimate of the charge state parameter based on multiple prior distribution values ​​of the charge state parameter, and the prior estimate of the covariance matrix of the charge state parameter, includes:

[0058] Based on the mean weight corresponding to the prior distribution value of each of the charged state parameters, the results obtained by processing the prior distribution values ​​of the charged state parameters through the charged filtering state equation are weighted to obtain the prior estimate value of the charged state parameter.

[0059] Based on the covariance weights corresponding to the prior distribution values ​​of each of the charged state parameters, the differences between the prior distribution values ​​of the charged state parameters and the prior estimated values ​​of the charged state parameters are weighted to obtain the prior estimated values ​​of the covariance matrix of the charged state parameters.

[0060] As a further feasible implementation of this application, updating the prior estimate of the charge state parameter when the prior estimate of the covariance matrix of the charge state parameter and / or the prior estimate of the charge state parameter satisfies a preset condition includes:

[0061] When the prior estimate of the covariance matrix of the charged state parameter is a pairwise matrix, a positive definite matrix, and / or the prior estimate of the covariance matrix of the charged state parameter is within a preset covariance boundary range, and / or the prior estimate of the charged state parameter is within a preset charged state boundary range, the prior estimate of the charged state parameter is updated.

[0062] As a further feasible implementation of this application, the prior estimated parameters based on capacity parameters and prior estimated values ​​of battery model parameters are used to perform posterior estimation on the prior estimated values ​​of the state of charge parameters to obtain updated values ​​of the state of charge parameters in the battery state parameters, including:

[0063] Based on the prior estimates of the charged state parameters and the prior estimates of the covariance matrix of the charged state parameters, multiple posterior distribution values ​​of the charged state parameters are obtained.

[0064] The multiple posterior distribution values ​​of the state of charge parameters, the prior estimated parameters of the capacity parameters, and the prior estimated values ​​of the battery model parameters are processed by the state of charge filtering measurement equation to obtain the third measurement mean and the third measurement covariance matrix.

[0065] Based on the third measurement mean and the third gain matrix corresponding to the third measurement covariance matrix, the prior estimates of the state of charge parameters and the prior estimates of the covariance matrix of the state of charge parameters are updated to obtain updated values ​​of the state of charge parameters. The updated values ​​of the state of charge parameters include the posterior estimates of the state of charge parameters and the posterior estimates of the covariance matrix of the state of charge parameters.

[0066] As a further feasible implementation of this application, the method further includes:

[0067] Based on the current information within a preset period, the charge and discharge state of the battery and the target polarity level corresponding to the charge and discharge state are determined; wherein, different polarity levels correspond to different correlation curves, and the correlation curves are the correlation curves between open circuit voltage and state of charge.

[0068] as well as

[0069] During the process of updating different battery state parameters based on different update frequencies, the battery state parameters are updated based on the target relationship curve corresponding to the target polarity level.

[0070] As a further feasible implementation of this application, the method further includes:

[0071] Determine the ratio of the change in open-circuit voltage to the change in state of charge within a preset period;

[0072] When the ratio is within a preset ratio range, the step of updating different battery state parameters based on different update frequencies is performed.

[0073] As a further feasible implementation of this application, the method further includes:

[0074] When the ratio is outside the preset ratio range, the predicted battery state is determined based on the current integral.

[0075] As a further feasible implementation of this application, the prediction of battery state based on the updated battery state parameters includes:

[0076] The maximum allowable discharge current within a preset discharge cycle is determined based on the updated battery state parameters.

[0077] As a further feasible implementation of this application, the method further includes:

[0078] During the final stage of charging, the battery's state-of-charge parameters are predicted using the following formula:

[0079]

[0080] Among them, SOC t U represents the state of charge after time t during charging, and SOC0 represents the state of charge at the initial moment of charging. t Ut is the measured voltage at time t during charging, and U0 is the measured voltage at the initial moment of charging. end K1 and K2 are preset system gain coefficients, where K1 is the cutoff voltage.

[0081] In a second aspect, this application provides an electronic device, including a processor; a memory for storing processor-executable instructions; wherein the processor is configured to perform the steps of the battery state prediction method described in any of the preceding claims.

[0082] Thirdly, this application provides a computer-readable storage medium having a computer program stored thereon, the computer program being loaded by a processor to perform the steps of the battery state prediction method described in any of the preceding claims.

[0083] Fourthly, this application provides a battery in which the battery state is determined by performing the steps of the battery state prediction method described in any of the preceding claims.

[0084] Fifthly, this application provides a vehicle including the battery as described above.

[0085] This application updates different battery state parameters based on different update frequencies, that is, by updating battery state parameters through different time scales. This can better adapt to the update needs of different battery state parameters, so that the updated battery state parameters can be used more accurately for updating battery state parameters. Attached Figure Description

[0086] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0087] To gain a more complete understanding of this application and its beneficial effects, the following description will be provided in conjunction with the accompanying drawings, wherein the same reference numerals in the following description denote the same parts.

[0088] Figure 1 This is a flowchart illustrating the steps of the battery state prediction method provided in the embodiments of this application;

[0089] Figure 2 This is a schematic flowchart illustrating a step for updating battery state parameters according to an embodiment of this application.

[0090] Figure 3 This is a schematic flowchart illustrating a step of filtering and updating capacity parameters according to an embodiment of this application.

[0091] Figure 4 This is a schematic diagram illustrating the specific steps for prior estimation of capacity parameters provided in an embodiment of this application.

[0092] Figure 5 This is a schematic flowchart illustrating a step for obtaining prior estimates of capacity parameters and their covariance matrices, as provided in an embodiment of this application.

[0093] Figure 6 A schematic flowchart illustrating the steps of a posterior estimation process for state of charge parameters provided in an embodiment of this application;

[0094] Figure 7 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0095] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the protection scope of this application.

[0096] To clearly understand the battery state prediction method provided in this application embodiment, the specific application scenarios of the battery state prediction method will be described below. Specifically, with the development of electric vehicles and hybrid vehicles, accurate prediction of the battery state of the vehicle battery can more accurately control the vehicle's power output, thereby effectively improving vehicle performance. However, battery state prediction relies on the joint determination of multiple battery state parameters, and as the battery is used, i.e., the battery health declines, various battery state parameters will also change. The related technology that updates various battery state parameters by using aging curves can no longer meet the needs of accurate battery state prediction.

[0097] Based on the above application scheme, the battery state prediction method provided in this application can better adapt to the update requirements of different battery state parameters by using different update frequencies to update different battery state parameters, so that the updated battery state parameters can be used more accurately for updating battery state parameters.

[0098] Furthermore, in other feasible embodiments, this application also provides a collaborative interaction architecture that takes battery parameters as input, outputs various battery state parameters in real time, and the various battery state parameters output are inputs to each other, thereby accelerating the correction rate of battery state parameters, improving the correction accuracy of battery state parameters, ensuring the power performance of the whole vehicle, and providing support for the strategies of other modules of the whole vehicle.

[0099] Furthermore, in other feasible embodiments, this application also provides an online identification and update of various battery state parameters based on the square root unscented Kalman filter algorithm, which can eliminate cell consistency and subsequent aging differences online, and improve the robustness of various battery state parameters.

[0100] The battery state prediction method provided in this application will be described in detail below.

[0101] For details, please refer to Figure 1 , Figure 1 This is a flowchart illustrating the steps of the battery state prediction method provided in the embodiments of this application, specifically including steps S110 to S120:

[0102] S110 collects the battery's output signal.

[0103] To achieve real-time updates of battery state parameters, the solution provided in this application embodiment will collect various output signals from the battery. These output signals typically include current signals, voltage signals, and charge signals, etc. These output signals will be used in conjunction with the current battery state parameters as output during subsequent battery state parameter updates, thereby achieving coordinated updates of various battery state parameters, accelerating the correction rate, and improving correction accuracy. For example, the open circuit voltage (OCV) in the voltage signal can typically be correlated with the state of charge (SOC) parameter in the battery state parameters through a preset curve, i.e., determined through the SOC-OCV curve.

[0104] S120 updates different battery state parameters based on different update frequencies, in order to predict the battery state based on the updated battery state parameters.

[0105] Based on the aforementioned acquisition of battery output signals, the solution provided in this application update different battery state parameters at different update frequencies based on these signals. Specifically, these battery state parameters typically include at least two of the following: state of charge (SCC) parameters, battery model parameters, and capacity parameters. Through the coordinated updating of at least two battery state parameters, these parameters serve as inputs to each other, enabling the battery state prediction model to update other battery state parameters. The mutual constraints and influences between different battery state parameters significantly accelerate the correction rate and improve the accuracy of battery state parameter correction. This ensures overall vehicle power performance while supporting strategies for other vehicle modules and adapts to the update requirements of different battery state parameters, effectively improving the update effect. Of course, in a feasible implementation, updating battery state parameters can simultaneously include updating SCC parameters, battery model parameters, and capacity parameters. For ease of understanding, specific update schemes for SCC parameters, battery model parameters, and capacity parameters will be described separately using different embodiments below.

[0106] Furthermore, the update frequency provided in this application can typically be described by the cumulative duration or change in charge during the battery charging and discharging process. In other words, updating different battery state parameters based on different update frequencies includes:

[0107] Different battery state parameters are updated when the cumulative duration and / or change in charge during the battery charging and discharging process meet different conditions.

[0108] Specifically, in the solution provided in this application, by statistically analyzing the cumulative duration and / or charge change during the battery charging and discharging process, triggering logic for different battery state parameters can be triggered separately, thereby executing the update process for different battery state parameters respectively.

[0109] For example, in a feasible implementation, updating different battery state parameters when the cumulative duration and / or change in charge during the battery charging and discharging process meet different conditions includes:

[0110] When the cumulative duration of battery charging and discharging reaches a first preset duration and / or the change in charge reaches a first preset change, the capacity parameter in the battery state parameters is updated; or

[0111] When the cumulative duration of battery charging and discharging reaches a second preset duration and / or the change in charge reaches a second preset change, the battery model parameters in the battery state parameters are updated; or

[0112] When the cumulative duration of the battery charging and discharging process reaches a third preset duration and / or the change in charge reaches a third preset change, the state of charge parameters in the battery state parameters are updated.

[0113] In other words, whenever the cumulative duration of the battery charging and discharging process reaches a first preset duration T1 or the cumulative discharge exceeds n1% of the state of charge (SOC), the capacity parameter in the battery state parameters, i.e., the maximum capacity of the battery, will be updated. Similarly, whenever the cumulative duration of the battery charging and discharging process reaches a first preset duration T2 or the cumulative discharge exceeds n2% of the state of charge (SOC), the battery model in the battery state parameters will be updated. The battery model can typically be used to characterize various battery characteristics. It can use simple resistors or various equivalent circuit models such as Thevenin's model. Furthermore, it can be used by adding RC modules (resistor-capacitor modules) or using more complex battery models such as DP (Dual Polarization) models that can more accurately represent battery characteristics. This application does not limit the specific type of battery model used. However, regardless of the model used, the battery model generally includes necessary coefficients related to impedance parameters to calculate the time constant τ and ohmic internal resistance R in the transient behavior of the battery. om and polarization resistance R p1 Information such as these is required, and the specific calculation process usually depends on the battery model used. In addition, the state of charge (SOC) is updated whenever the cumulative duration of the battery charging and discharging process reaches a first preset duration T3 or the cumulative discharge exceeds n3% of the SOC.

[0114] Specifically, the battery capacity parameters change relatively slowly, while the battery model parameters change more rapidly with battery use. The state of charge (SOC) parameter has high requirements for both the rate of change and accuracy, requiring high-frequency calculations. Therefore, the battery capacity parameters can be defined using a first-order time scale, the battery model parameters using a second-order time scale, and the SOC parameter using a third-order time scale. That is, the first preset duration T1 is greater than the second preset duration T2, the second preset duration T2 is greater than the third preset duration T3, the first preset change amount n1% is greater than the second preset change amount n2%, and the second preset change amount n2% is greater than the third preset change amount n3. For example, in a feasible implementation, the first preset duration T1 can be set to one hour, the second preset duration T2 can be set to one minute, and the third preset duration can be set to the sampling period, that is, after each sampling of battery information, the state of charge parameter SOC will be updated once. Alternatively, the first preset change amount n1% can be set to 60%, the second preset change amount n2% can be set to 20%, and the third preset change amount n3% can be set to 0.1%, that is, whenever the state of charge parameter changes, the state of charge parameter SOC will be updated again.

[0115] Of course, the aforementioned implementation scheme is merely one feasible solution and should not be construed as a limitation of this application. In fact, based on the different battery state parameter update requirements, using different time scales, i.e., different update frequencies, to update different battery state parameters should all be within the scope of protection claimed in this application. Furthermore, it should be noted that after each update of the current state parameter, the accumulated duration or charge change of the corresponding current state parameter is clearly recorded and recounted. The embodiments of this application will not be elaborated upon here.

[0116] Building upon the aforementioned solutions, this application further provides a collaborative interaction architecture for real-time estimation of different current state parameters. The estimated current state parameters also serve as additional input parameters to update other current state parameters, making the accuracy of different current state parameters interdependent, thereby accelerating the correction rate and improving correction accuracy. In particular, by predicting deviations in both the prior and posterior stages, the collaborative correction effect can be effectively improved. For details, please refer to [link to relevant documentation]. Figure 2 , Figure 2 This is a flowchart illustrating a step for updating battery state parameters according to an embodiment of this application, specifically including steps S210 to S230:

[0117] S210, based on the capacity parameters before the update and the prior estimate of the state of charge parameters, the capacity parameters in the battery state parameters are updated.

[0118] In this embodiment, updating the capacity parameter in the battery state parameters depends on the prior estimate of the state of charge (SOC) parameter. Specifically, the capacity parameter can be updated through a filtering algorithm, such as the capacitive Kalman filter, the H-infinity filter, etc. However, as a feasible embodiment of this application, a square root-based unscented Kalman filter algorithm is provided, which can effectively eliminate the Taylor expansion bias of second-order and higher nonlinear functions, thereby effectively improving the correction accuracy.

[0119] S220, based on the battery model parameters before the update and the prior estimates of the state of charge parameters, update the battery model parameters in the battery state parameters.

[0120] Similar to the aforementioned implementation scheme for updating capacity parameters, the technical solution provided in this application embodiment will update the battery model parameters by using prior estimates of battery model parameters and state of charge (SOC) parameters, and by using a square root-based unscented Kalman filter algorithm to achieve prior and posterior estimates of battery model parameters.

[0121] S230, based on the state of charge parameters before the update, as well as the prior estimated parameters of the capacity parameters and the prior estimated values ​​of the battery model parameters, the state of charge parameters in the battery state parameters are updated.

[0122] In this embodiment, by using a collaborative interaction architecture, the estimated values ​​of different battery state parameters are additionally used to update other battery state parameters. Therefore, for updating the state of charge (SOC) parameter among the battery state parameters, it is often necessary to process the SOC parameter before updating based on the prior estimated parameters of the capacity parameter and the prior estimated values ​​of the battery model parameters. Specifically, similar to the above, the update of the SOC parameter can also be implemented based on the square root unscented Kalman filter algorithm.

[0123] The collaborative interaction architecture provided in this application, where each battery state parameter depends on the estimated value of other parameters, makes the accuracy of different current state parameters interdependent. This accelerates the correction rate while improving the accuracy of parameter correction, thereby effectively improving subsequent prediction and estimation of the battery state. The specific processing of different parameters in the battery state parameters, such as capacity parameters, battery model parameters, and state of charge (SOC) parameters, using the square root-based unscented Kalman filter algorithm, will be detailed in the implementation scheme provided later. Please refer to the following description for details.

[0124] First, in order to update different battery state parameters, in this embodiment, corresponding system state equations and measurement equations are constructed for the filtering system of each battery state parameter, and the state matrix of each system is also defined. Details are as follows.

[0125] (1) State equations and measurement equations of the battery model parameter filtering system

[0126] Within the filtering system of the battery model parameters, the following data matrix Φ and state matrix θ are defined, as follows:

[0127]

[0128] Where, Φ k That is, the data matrix at time k, U oc，k-1 This represents the OCV (open-circuit voltage) lookup value at time k-1, which is the open-circuit voltage obtained by looking up the preset SOC-OCV mapping curve through the battery's state of charge parameter SOC. U term，k-1 This represents the lowest voltage of a single cell in the battery pack at time k-1, I. k and I k-1 Let θ be the current at time k and time k-1, respectively. k This is the state matrix at time k, which contains coefficients related to the impedance parameters, namely θ1, θ2, and θ3, used to calculate the parameters in the battery model. The specific calculation details will be provided in the subsequent content.

[0129] Based on the aforementioned data matrix and state matrix, the state equations and measurement equations within the filtering system of the battery model parameters are as follows:

[0130]

[0131] Wherein, the state equation θ k+1 =θ k This indicates that the state matrix within the filtering system remains unchanged, while the measurement equation Z para，k =Φ k ×θ k Represents dynamic voltage Z para，k The difference between the open-circuit voltage (OCV) and the terminal voltage is the product of the data matrix and the state matrix, used to achieve a posteriori estimation of the battery model parameters.

[0132] (2) State equations and measurement equations of the SOC filter system for charge state parameters

[0133] In a filter system with state-of-charge (SOC) parameters, the state matrix can be defined as:

[0134] X k =[SOC k U p1，k ] T

[0135] Among them, X k Let SOC be the state matrix of the SOC filtering system at time k, consisting of two-dimensional components SOC. k and U p1，k Composition, in which SOC k Let U be the state-of-charge (SOC) parameter at time k. p1，k Let be the polarization voltage value at time k. The polarization voltage refers to the deviation of the electrode potential from its equilibrium potential caused by current flow; in simpler terms, it describes the transient response behavior of the battery under varying load. At this time, the state equation of the SOC filter system is calculated as follows:

[0136]

[0137] Among them, X k (1, 1) represents the data in the first row and first column of the state matrix, which is also known as the SOC. k That is, the charged state parameters in the state matrix of the SOC filter system at time k are derived from the charged state parameters X at time k-1. k-1 (1, 1), which is also known as SOC k-1 Subtract the charge Q that changes during the time interval. delta The ratio to the capacity parameter Cn is obtained, while X k (2,1) represents the data in the second row and first column of the state matrix, i.e., U p1，k That is, the polarization voltage value in the state matrix of the SOC filter system at time k is the polarization voltage value X at time k-1. k-1 (2, 1), that is, U p1，k-1 The time constant τ1 and the data sampling period T at time k are used. delta，k Based on a certain ratio, increase the current at time k and the polarity resistance R at time k. p1 and the time constant τ1 and the data sampling period T at time k. delta，k The product of the ratios between them is obtained, where the time constant τ1 and the polarity resistance R are also included. p1 It is usually related to the circuit model parameters, which can typically be obtained by filtering the state matrix θ of the battery model parameters. The specific calculation process will also be provided in the corresponding content later.

[0138] Furthermore, the measurement equations for the SOC filter system are as follows:

[0139] Y k =Uoc -I k ×R om -X k (2,1)

[0140] Among them, Y k Let U be the model predicted voltage at time k. oc It is the open-circuit voltage, R om For the internal resistance of ohms, X k (2,1) represents the polarization voltage value in the state matrix of the SOC filter system at time k, where R om It is usually obtained from the state matrix θ of the battery model parameter filtering system.

[0141] (3) State equations and measurement equations in a filter system with capacity parameters

[0142] Among them, the state matrix Cn of the capacity parameter has only one dimension, namely the maximum battery capacity, and the corresponding state equation of the capacity parameter filtering system is as follows:

[0143] Cn k+1 =Cn k

[0144] This means that the state matrix within the filter system with the capacity parameter remains unchanged, while the measurement equations within the filter system with the capacity parameter are consistent with the measurement equations of the SOC filter system, both being:

[0145] Y k =U oc -I k ×R om -X k (2,1)

[0146] The description of the measurement equations for the capacity parameter within the filter system will not be elaborated here.

[0147] Of course, in addition to the state equations and measurement equations constructed from the different battery state parameters provided above, in order to implement the unscented Kalman filter algorithm provided in this application embodiment, the unscented transform weight parameters used in the unscented Kalman filter algorithm will be described in detail below.

[0148] Specifically, as described above, the state vector dimension L of the SOC filtering system is... sta =2, i.e., SOC value and polarization voltage value, and the state vector dimension L of the battery model parameter filtering system. para =3, i.e., θ1, θ2, θ3, the dimension L of the state vector of the capacity filter system. Cn=1. Furthermore, the unscented transform weight parameters are usually also related to the sigma point spread range constant α and the prior information parameter β, where α and β are typically preset values. In this case, the weights used in approximating the posterior probability distribution of the output variable through the unscented transform include the mean weight ω. m Covariance weight ω c The specific calculation process is as follows:

[0149]

[0150] The number of sigma points obtained through unscented transformation varies depending on the state vector dimension of different filtering systems, typically around 2L+1 points. For example, in a SOC filtering system, the number of sigma points obtained through unscented transformation is 2×2+1, or 5 points. For a battery model parameter filtering system, the number of sigma points is 2×3+1, or 7 points. Except for the first point, which is the mean, the other points are symmetrically distributed on both sides of the mean. Therefore, for the first point, the mean weight... For the remaining points, their mean weights

[0151] The corresponding calculation process for the covariance weights is as follows:

[0152]

[0153] Similarly, for the first point, its covariance weights For the remaining points, their covariance weights The L, α, and β used in the weight calculation process have been provided above and will not be repeated here in this embodiment.

[0154] In addition, to achieve filtering and updating of various battery state parameters, it is usually necessary to define various initialization parameters. These initialization parameters typically include the state vector of the SOC filtering system (initial SOC value and initial polarization voltage), the state vector of the battery model parameter filtering system, and the initial capacity value of the battery (pack). Furthermore, there are P (state estimation error covariance matrix), Q (process noise covariance matrix), and R (measurement noise covariance matrix) corresponding to each parameter filtering system. Among these, Q and R are usually predefined values ​​and do not participate in the updating during Kalman filtering, while P usually needs to be updated.

[0155] Building upon the aforementioned foundation, the following sections will provide a detailed explanation of the update process for different parameters in the battery state parameters. For the sake of simplicity, the explanation will focus primarily on the update process of the capacity parameter. For the update processes of other parameters, necessary formulas will be provided to assist in the explanation, and repeated definitions will not be elaborated upon.

[0156] For details, please refer to [link / reference]. Figure 3 , Figure 3 This is a flowchart illustrating a step for filtering and updating capacity parameters according to an embodiment of this application. Specifically, it includes steps S310 to S320:

[0157] S310, the capacity parameter before the update is estimated using a preset filtering algorithm to obtain the prior estimate value of the capacity parameter.

[0158] In the embodiments of this application, updating battery state parameters, such as updating capacity parameters, usually involves prior estimation and posterior estimation of capacity parameters. That is, in the prior process, the capacity parameters before the update, that is, the estimated value of the capacity parameters in the previous sampling period, are used in combination with the Kalman filter algorithm of unscented transform to realize the prior estimation of capacity parameters and obtain the prior estimated value of capacity parameters.

[0159] S320, based on the prior estimate of the state of charge parameter, perform a posteriori estimation on the prior estimate of the capacity parameter to obtain the updated value of the capacity parameter in the battery state parameters.

[0160] Based on the prior estimate of the capacity parameter, this embodiment of the application further performs a posterior estimation of the prior estimate of the capacity parameter based on the prior estimate of the state of charge parameter (SOC), thereby obtaining the updated value of the capacity parameter in the battery state parameters. At this time, the obtained updated value can be further regarded as the estimated value of the capacity parameter in the next processing cycle, and used to participate in the calculation of the prior estimate of the capacity parameter in the next processing cycle. Based on this, the update of the capacity parameter is completed cyclically.

[0161] For reasons similar to those mentioned above, it can be understood that in another embodiment, the step of filtering and updating the battery model parameters, namely step S220, typically includes:

[0162] The battery model parameters before the update are estimated using a preset filtering algorithm to obtain the prior estimate values ​​of the battery model parameters.

[0163] Based on the prior estimates of the state of charge parameters, the prior estimates of the battery model parameters are posteriorly estimated to obtain the updated values ​​of the battery model parameters in the battery state parameters.

[0164] Furthermore, in another embodiment, the step of filtering and updating the state of charge parameters, namely step S230, typically includes:

[0165] The prior estimate of the charge state parameters before the update is obtained by using a preset filtering algorithm.

[0166] Based on the prior estimated parameters of the capacity parameter and the prior estimated values ​​of the battery model parameters, the prior estimated values ​​of the state of charge parameter are posteriorly estimated to obtain the updated values ​​of the state of charge parameter in the battery state parameters.

[0167] While the data processed may differ slightly, the underlying approaches to updating these different battery state parameters are largely the same, including the filtering algorithms and prior and subsequent estimation methods. Those skilled in the art can determine the specific update process for each battery state parameter based on the foregoing description and the detailed implementation information provided below.

[0168] Building upon the aforementioned procedures for prior and posterior estimation of capacity parameters, the following section will explain the specific process of prior estimation. For details, please refer to [link to relevant documentation]. Figure 4 , Figure 4 This application provides a schematic flowchart illustrating the specific steps for prior estimation of capacity parameters, specifically including steps S410 to S430:

[0169] S410, based on the covariance matrix of the capacity parameter before the update, obtain multiple prior distribution values ​​of the capacity parameter.

[0170] In the solution provided in this application embodiment, the covariance matrix of the capacity parameter before the update is used, which is the state estimation error covariance matrix of the capacity parameter in the filtering system. The capacity parameter decomposition matrix S is obtained using Cholesky decomposition, i.e. in This is the posterior estimate of the covariance matrix of the capacity parameter at time k-1, which is the previous time step. The superscript + indicates the estimated value determined during the posterior estimation process (the same applies thereafter, unless otherwise specified). Then, based on the unscented transformation and the previously obtained decomposition matrix S, multiple prior distribution values ​​of the capacity parameter can be obtained:

[0171]

[0172] This can be understood as the first prior distribution value Cn1 among multiple prior distribution values ​​being the weighted average of the posterior distribution values ​​of the multiple capacity parameters determined at the previous time step. The remaining 2L prior distribution values ​​are paired up (Cn) i Cn 2L+3-i ), where i takes values ​​between 2 and L+1. For the capacity parameter, L is usually the dimension of the aforementioned capacity parameter, which is 1, and they are symmetrically distributed on the weighted mean. On both sides, among which the weighted mean The offset is This is typically associated with the decomposition matrix S, the i-th prior distribution value, and the constant term λ. Introducing the square root to generate the sigma point set, i.e., multiple prior distribution values, effectively improves numerical stability, reduces the complexity of calculating the covariance matrix, better preserves the positive definiteness of the covariance matrix, and more accurately reflects the true shape of the state distribution, especially under nonlinear transformations, thus helping to improve the estimation accuracy of the filter.

[0173] For reasons similar to those mentioned above, prior estimation of battery model parameters and SOC parameters can typically employ similar processing steps. The difference lies in the need to use the covariance matrix of the battery model parameters and SOC parameters before the update to generate prior distribution values. In other words, the aforementioned steps use a preset filtering algorithm to perform prior estimation of the battery model parameters before the update, obtaining the prior estimated values ​​of the battery model parameters. Specifically, this includes:

[0174] Based on the covariance matrix of the battery model parameters before the update, multiple prior distribution values ​​of the battery model parameters are obtained.

[0175] Specifically, by using the covariance matrix of the battery model parameters before the update, which is also the state estimation error covariance matrix of the battery model parameters in the filtering system, The Cholesky decomposition is used to obtain the decomposition matrix S′ of the battery model parameters, i.e. in This is the posterior estimate of the covariance matrix of the battery model parameters at time k-1, which is the previous time step. Then, based on the unscented transformation and the previously obtained decomposition matrix S′, multiple prior distribution values ​​of the battery model parameters can be obtained:

[0176]

[0177] Among the multiple prior distribution values ​​of the battery model parameters, the first prior distribution value θn1 is taken as the weighted mean of the posterior distribution values ​​of the multiple battery model parameters determined at the previous time step. The remaining 2L prior distribution values ​​are paired up (θn) i ,θn 2L+3-i), where i takes values ​​between 2 and L+1. For battery model parameters, L is usually the dimension of the battery model parameters mentioned above, which is 3, and they are symmetrically distributed on the weighted mean. The definitions of the other parameters can be obtained by referring to the definitions of multiple prior distribution values ​​in the capacity parameters, which will not be repeated here in the embodiments of this application.

[0178] Similarly, based on a similar implementation scheme, the aforementioned steps perform prior estimation of the charge state parameters before the update using a preset filtering algorithm to obtain the prior estimate values ​​of the charge state parameters, specifically including:

[0179] Based on the covariance matrix of the charge state parameters before the update, multiple prior distribution values ​​of the charge state parameters are obtained.

[0180] Specifically, this involves using the covariance matrix of the state-of-charge (SOC) parameters before the update, which is also the state estimation error covariance matrix in the SOC filtering system. The Cholesky decomposition is used to obtain the decomposition matrix S″ of the charged state parameter SOC, i.e. in This is the posterior estimate of the covariance matrix of the charged state parameter SOC at time k-1, which is the previous time. Then, based on the unscented transformation and the previously obtained decomposition matrix S", multiple prior distribution values ​​of the charged state parameter can be obtained:

[0181]

[0182] Among the multiple prior distribution values ​​of the charge state parameter SOC, the first prior distribution value Xn1 is taken as the weighted average of the multiple posterior distribution values ​​of the charge state parameters determined at the previous time step. The remaining 2L prior distribution values ​​are paired up (Xn) i , Xn 2L+3-i ), where i takes values ​​between 2 and L+1. For the state-of-charge (SOC) parameter, L is usually the dimension of the SOC state vector, which is 2, and the vectors are symmetrically distributed on the weighted mean. The definitions of the other parameters can be obtained by referring to the definitions of multiple prior distribution values ​​in the capacity parameters, which will not be repeated here in the embodiments of this application.

[0183] S420, based on multiple prior distribution values ​​of the capacity parameter, obtain the prior estimate of the capacity parameter and the prior estimate of the covariance matrix of the capacity parameter.

[0184] Given the aforementioned multiple prior distribution values ​​of the parameters obtained through unscented transformation, prior estimates of the capacity parameters and the covariance matrix of the capacity parameters can be further predicted based on the state equations within the corresponding parameter system.

[0185] Of course, for similar reasons, the prior estimation process for battery model parameters and SOC parameters also involves calculating the battery model parameters and their prior estimates, as well as the SOC parameters and their prior estimates. For example, in another embodiment, in the prior estimation process for battery model parameters, the step of performing prior estimation on the battery model parameters before the update using a preset filtering algorithm to obtain the prior estimates of the battery model parameters further includes:

[0186] Based on multiple prior distribution values ​​of the battery model parameters, prior estimates of the battery model parameters and prior estimates of the covariance matrix of the battery model parameters are obtained.

[0187] Alternatively, in yet another embodiment, the step of performing a priori estimation of the charge state parameters before the update using a preset filtering algorithm to obtain a priori estimated values ​​of the charge state parameters further includes:

[0188] Based on multiple prior distribution values ​​of the charged state parameters, prior estimates of the charged state parameters and prior estimates of the covariance matrix of the charged state parameters are obtained.

[0189] To clearly understand the calculation of the prior estimates of each battery state parameter and its covariance matrix, please refer to [link to relevant documentation]. Figure 5 , Figure 5 Taking the capacity parameter as an example, this paper provides a flowchart illustrating the steps to obtain prior estimates of the capacity parameter and its covariance matrix. Specifically, it includes the following steps:

[0190] S510, based on the mean weight corresponding to the prior distribution value of each capacity parameter, the results obtained by processing the prior distribution value of the capacity parameter through the capacity filtering state equation are weighted to obtain the prior estimate value of the capacity parameter.

[0191] In this embodiment, the prior estimate of the capacity parameter can be obtained using the mean weight corresponding to the prior distribution value of each capacity parameter, i.e., the aforementioned mean weight ω. m This is used to weight the results obtained by processing each prior distribution value through the capacity filter state equation, that is, the prior estimate of the capacity parameter. The calculation process is as follows:

[0192]

[0193] Here, the function f() represents the state equation within the filter system for the capacity parameter. For example, as mentioned earlier, in the prior estimation of the capacity parameter, the function f() used is the state equation Cn. k+1 =Cn k Therefore, f(Cn) i This means that for each prior distribution value Cn i Based on the results obtained through the aforementioned function, and combined with each prior distribution value Cn i The corresponding mean weight ω m By performing a weighted summation, we can obtain the prior estimate of the capacity parameter. Wherein, each prior distribution value Cn i The corresponding mean weight ω m This has also been specifically provided above. In addition, the superscript "-" corresponds to the superscript "+", representing the prior estimate determined during the prior estimation process (the same applies thereafter, unless otherwise specified).

[0194] S520, based on the covariance weights corresponding to the prior distribution values ​​of each capacity parameter, the difference between the prior distribution values ​​of the capacity parameter and the prior estimated values ​​of the capacity parameter is weighted to obtain the prior estimated values ​​of the covariance matrix of the capacity parameter.

[0195] Furthermore, regarding the prior estimates of the capacity parameters obtained above... Based on this, the solution provided in the embodiments of this application can further improve the prior estimate of the covariance matrix of the capacity parameter. The update is performed, whereby the prior estimate of the covariance matrix of the capacity parameter is... The specific calculation process is as follows:

[0196]

[0197] Where, ω c，i For each prior distribution value Cn i The specific setting process for the corresponding covariance weights has been provided above. The update depends on the prior distribution value Cn of each capacity parameter using covariance weights. i Prior estimates of capacity parameters The difference matrix between the two and its corresponding transpose are weighted and then summed with the process noise covariance matrix under the capacity parameter filtering system.

[0198] Based on a similar description as described above, the prior estimates of the battery model parameters and their covariance matrix can be obtained in the following way: that is, in another embodiment, obtaining the prior estimates of the battery model parameters and the prior estimates of the covariance matrix of the battery model parameters based on multiple prior distribution values ​​of the battery model parameters includes:

[0199] Based on the mean weight corresponding to the prior distribution value of each battery model parameter, the results obtained by processing the prior distribution values ​​of the battery model parameters through the battery model filtering state equation are weighted to obtain the prior estimate value of the battery model parameter.

[0200] Based on the covariance weights corresponding to the prior distribution values ​​of each battery model parameter, the differences between the prior distribution values ​​and the prior estimates of the battery model parameters are weighted to obtain the prior estimates of the covariance matrix of the battery model parameters.

[0201] Of course, it should be noted that the prior estimates obtained during the estimation of battery model parameters will affect some parameters within the battery model. These parameters will participate in the subsequent prior estimates of the state of charge (POC) parameters. For example, the prior estimates of the POC parameters require the use of the state equations of the filtered POC system, which involve the time constant τ1 and the ohmic internal resistance R. om and polarity resistor R p1 In this embodiment, the specific calculation formula for the above parameters is obtained through the prior estimate of the battery model parameters as follows:

[0202]

[0203] in, This represents the value in the first row and first column of the prior estimates of the battery model parameters at time k, which is the current time. It can also be understood as the prior estimate of θ1 among θ1, θ2, and θ3 at time k. Similarly, and Let θ2 and θ3 represent the prior estimates of θ2 and θ3 at time k, respectively. All other parameters have been described above and will not be repeated here in the embodiments of this application.

[0204] Furthermore, in another embodiment, obtaining the prior estimate of the charge state parameter based on multiple prior distribution values ​​of the charge state parameter, and the prior estimate of the covariance matrix of the charge state parameter, includes:

[0205] Based on the mean weight corresponding to the prior distribution value of each of the charged state parameters, the results obtained by processing the prior distribution values ​​of the charged state parameters through the charged filtering state equation are weighted to obtain the prior estimate value of the charged state parameter.

[0206] Based on the covariance weights corresponding to the prior distribution values ​​of each of the charged state parameters, the differences between the prior distribution values ​​of the charged state parameters and the prior estimated values ​​of the charged state parameters are weighted to obtain the prior estimated values ​​of the covariance matrix of the charged state parameters.

[0207] S430, if the prior estimate of the covariance matrix of the capacity parameter and / or the prior estimate of the capacity parameter meet the preset conditions, update the prior estimate of the capacity parameter.

[0208] In this embodiment, after obtaining prior estimates of the battery state parameters and their covariance matrix using the aforementioned method, it is possible to further determine whether the obtained prior estimates meet certain requirements, thereby determining whether the estimated values ​​of the battery state parameters need to be updated. For example, if the prior estimates of the covariance matrix of the capacity parameter and / or the prior estimates of the capacity parameter meet preset conditions, the prior estimates of the capacity parameter are updated. Alternatively, in another embodiment, during the prior estimation of the battery model parameters, the prior estimation of the battery model parameters before the update using a preset filtering algorithm to obtain the prior estimates of the battery model parameters includes:

[0209] If the prior estimate of the covariance matrix of the battery model parameters and / or the prior estimate of the battery model parameters meets a preset condition, the prior estimate of the battery model parameters is updated.

[0210] Furthermore, in another embodiment, during the prior estimation of the state of charge parameters, the prior estimation of the state of charge parameters before the update using a preset filtering algorithm to obtain the prior estimated values ​​of the state of charge parameters includes:

[0211] If the prior estimate of the covariance matrix of the charge state parameter and / or the prior estimate of the charge state parameter meets a preset condition, the prior estimate of the charge state parameter is updated.

[0212] Specifically, to facilitate understanding of the judgment conditions provided in the embodiments of this application for determining the need to update the prior estimate, the following will take the capacity parameter as an example. Specifically, step S430 includes:

[0213] When the prior estimate of the covariance matrix of the capacity parameter is a pairwise matrix, a positive definite matrix, and / or the prior estimate of the covariance matrix of the capacity parameter is within a preset covariance boundary range, and / or the prior estimate of the capacity parameter is within a preset capacity boundary range, the prior estimate of the capacity parameter is updated.

[0214] Specifically, the prior estimate of the covariance matrix of the capacity parameter... For the prior estimates of the covariance matrix of the pairwise matrix, the positive definite matrix, and / or the capacity parameter. Located within the preset covariance boundary range P1 to P2, and / or, the prior estimate of the capacity parameter. If the value is within the preset capacity boundary range C1 to C2, then the updated prior estimates of the capacity parameters and their covariance matrix can be considered relatively accurate. In this case, the current prior estimates can be used as the latest estimates of the capacity parameters for posterior estimation of the capacity parameters or for updating other battery state parameters. Otherwise, the current prior estimates of the capacity parameters and their covariance matrix can be considered to have some deviation or not in line with the actual situation. In this case, the posterior estimates obtained at the previous moment can still be used as estimates of the capacity parameters for posterior estimation of the capacity parameters or for updating other battery state parameters.

[0215] Of course, based on a similar principle, in another embodiment, updating the prior estimate of the battery model parameters when the prior estimate of the covariance matrix of the battery model parameters and / or the prior estimate of the battery model parameters meets a preset condition includes:

[0216] When the prior estimate of the covariance matrix of the battery model parameters is a pairwise matrix, a positive definite matrix, and / or the prior estimate of the covariance matrix of the battery model parameters is within a preset covariance boundary range, and / or the prior estimate of the battery model parameters is within a preset model parameter boundary range, the prior estimate of the battery model parameters is updated.

[0217] Alternatively, in yet another embodiment, updating the prior estimate of the charge state parameter when the prior estimate of the covariance matrix of the charge state parameter and / or the prior estimate of the charge state parameter satisfies a preset condition includes:

[0218] When the prior estimate of the covariance matrix of the charged state parameter is a pairwise matrix, a positive definite matrix, and / or the prior estimate of the covariance matrix of the charged state parameter is within a preset covariance boundary range, and / or the prior estimate of the charged state parameter is within a preset charged state boundary range, the prior estimate of the charged state parameter is updated.

[0219] Among them, the preset covariance boundary range, the preset model parameter boundary range, and the preset charge state boundary range can all be understood as some ranges set in advance, and will not be repeated in the embodiments of this application.

[0220] After completing the aforementioned prior estimates of capacity parameters and other battery state parameters, it is understood that a corresponding posterior estimation process needs to be executed. That is, after each prior estimate of the capacity parameter is completed, the prior estimate of the capacity parameter will be combined with the estimates of other battery state parameters to further complete the posterior estimate of the capacity parameter, so as to obtain a more accurate estimate for the next round of parameter updates. Similarly, after completing a prior estimate of the battery model parameters, the prior estimate of the battery model parameters will be combined with the estimates of other battery state parameters to further complete the posterior estimate of the battery model parameters, so as to obtain a more accurate estimate for the next round of parameter updates.

[0221] Furthermore, it should be noted that, unlike the prior estimation process where the prior estimates of capacity parameters and battery model parameters affect the prior estimates of state-of-charge (POC) parameters, in the posterior estimation process, the posterior estimates of POC parameters, in turn, affect the posterior estimates of capacity parameters and battery model parameters. Therefore, to more clearly understand the solution provided in the embodiments of this application, the posterior estimation process will be illustrated using the posterior estimation process of POC parameters as an example. For details, please refer to [link to relevant documentation]. Figure 6 , Figure 6 This application provides a flowchart illustrating the steps of a posterior estimation process for state-of-charge parameters, specifically including steps S610 to S630:

[0222] S610, based on the prior estimates of the charged state parameters and the prior estimates of the covariance matrix of the charged state parameters, multiple posterior distribution values ​​of the charged state parameters are obtained.

[0223] Similar to the aforementioned prior estimation process that utilizes the posterior estimate from the previous time step to obtain the prior distribution value, this embodiment also uses an unscented transformation to obtain multiple posterior distribution values ​​of the charged state parameters. That is, it is necessary to first obtain the prior estimate of the covariance matrix of the charged state parameters through Cholesky decomposition. Then, based on the unscented transformation and the aforementioned decomposition matrix, multiple posterior distribution values ​​of the charged state parameters can be obtained. Specifically, the obtained multiple posterior distribution values ​​are as follows:

[0224]

[0225] Among them, the first posterior distribution value X1 of the multiple posterior distribution values ​​of the charged state parameter can be the prior estimate value of the charged state parameter calculated in the prior estimation process at time k, which is the current time. The remaining 2L posterior distribution values ​​are symmetrically distributed in pairs on the prior estimates. Both sides, and the definitions of the remaining parameters can be referred to the description of the prior process, which will not be repeated here in the embodiments of this application.

[0226] Furthermore, based on a similar approach, in another embodiment, during the posterior estimation of the battery model parameters, the step of performing posterior estimation on the prior estimated values ​​of the battery model parameters based on the prior estimated values ​​of the state of charge parameters to obtain updated values ​​of the battery model parameters in the battery state parameters includes:

[0227] Based on the prior estimates of the battery model parameters and the prior estimates of the covariance matrix of the battery model parameters, multiple posterior distribution values ​​of the battery model parameters are obtained.

[0228] Alternatively, in yet another embodiment, during the posterior estimation of the capacity parameter, the prior estimate of the capacity parameter based on the prior estimate of the state of charge parameter is used to perform a posterior estimation to obtain an updated value of the capacity parameter in the battery state parameters, including:

[0229] Based on the prior estimates of the capacity parameter and the prior estimates of the covariance matrix of the capacity parameter, multiple posterior distribution values ​​of the capacity parameter are obtained.

[0230] The specific implementation process for obtaining the posterior distribution values ​​of different battery state parameters will not be repeated here. It is only necessary to replace the prior estimates of the covariance matrix of the corresponding processed state of charge parameters and the prior estimates of the state of charge parameters with the prior estimates of the covariance matrix of the battery model parameters and the prior estimates of the battery model parameters, or with the prior estimates of the covariance matrix of the capacity parameters and the prior estimates of the capacity parameters, respectively.

[0231] S620 processes the multiple posterior distribution values ​​of the state of charge parameters, the prior estimated parameters of the capacity parameters, and the prior estimated values ​​of the battery model parameters through the state of charge filtering measurement equation to obtain the third measurement mean and the third measurement covariance matrix.

[0232] Based on the aforementioned scheme, by inputting multiple posterior distribution values ​​of the charge state parameters into the measurement equation of the charge state parameter filtering system, i.e., the aforementioned Y... k =U oc -I k×R om -X k In (2,1), the measurement results of each posterior distribution value can be obtained. Then, the mean weight of each posterior distribution value can be used to weight this part of the measurement results, so as to obtain the measurement mean Y of the posterior distribution of the charge state parameter. k That is, the third measurement mean, specifically, the measurement mean of the state of charge parameter Y. k The calculation process is as follows:

[0233] Y u，i =u(X) i )

[0234]

[0235] Where u() is the aforementioned measurement equation, ω m，i It is the mean weight.

[0236] Furthermore, by combining the measured mean of the charge state parameters mentioned above, and the covariance weights provided earlier, the measurement error covariance matrix of the charge state parameters can be obtained. The specific calculation formula is as follows:

[0237]

[0238] Where, ω c，i R represents the covariance weights, while R is the measurement noise covariance matrix under the charged state parameter filtering system.

[0239] Furthermore, the covariance P can be obtained by further weighted summation. xy′

[0240]

[0241] Each parameter in the above formula has been provided in the foregoing description, and will not be redefined here.

[0242] Of course, based on a similar processing procedure, in another embodiment, the posterior distribution estimation of the battery model parameters also involves the calculation of the measurement mean and measurement covariance matrix of the battery model parameters. That is, the step of performing posterior estimation on the prior estimated values ​​of the battery model parameters based on the prior estimated values ​​of the state of charge parameters to obtain the updated values ​​of the battery model parameters in the battery state parameters includes:

[0243] The battery model filtering measurement equation is used to process multiple posterior distribution values ​​of the battery model parameters and prior estimates of the state of charge parameters to obtain the second measurement mean and the second measurement covariance matrix.

[0244] It should be noted that in the process of calculating the measurement mean and measurement covariance matrix of the battery model parameters, the measurement equation used is the measurement equation within the filtering system of the battery model parameters. At this time, the data processed corresponds to multiple posterior distribution values ​​of the battery model parameters. Since the measurement equation within the filtering system of the battery model parameters involves OCV lookup table values, i.e., OCV values ​​determined by the SOC parameters, the SOC parameters used here are the SOC values ​​in the prior estimates of the state of charge parameters obtained above. That is, the measurement mean and measurement covariance matrix of the battery model parameters, i.e., the calculation of the second measurement mean and the second measurement covariance matrix depends on the prior estimates of the state of charge parameters.

[0245] Furthermore, in another embodiment, the posterior distribution estimation of the capacity parameter also involves the calculation of the measured mean and measurement covariance matrix of the capacity parameter. That is, based on the prior estimate of the state of charge parameter, the prior estimate of the capacity parameter is posteriorly estimated to obtain the updated value of the capacity parameter in the battery state parameters, including...

[0246] Based on the first measurement mean and the first gain matrix corresponding to the first measurement covariance matrix, the prior estimate of the capacity parameter and the prior estimate of the covariance matrix of the capacity parameter are updated to obtain the updated value of the capacity parameter. The updated value of the capacity parameter includes the posterior estimate of the capacity parameter and the posterior estimate of the covariance matrix of the capacity parameter.

[0247] The measurement equations used here are those within the capacity parameter filtering system (usually the same as those within the state-of-charge parameter filtering system), and these also involve the OCV lookup table value U. oc That is, the OCV value determined by the SOC parameter. Here, the SOC parameter used is the same as that mentioned above, which is the SOC value in the prior estimate of the charge state parameter obtained above, that is, the measurement mean and measurement covariance matrix of the capacity parameter. In other words, the calculation of the first measurement mean and the first measurement covariance matrix depends on the prior estimate of the charge state parameter.

[0248] S630, based on the third measurement mean and the third gain matrix corresponding to the third measurement covariance matrix, the prior estimate of the state of charge parameter and the prior estimate of the covariance matrix of the state of charge parameter are updated to obtain the updated value of the state of charge parameter. The updated value of the state of charge parameter includes the posterior estimate of the state of charge parameter and the posterior estimate of the covariance matrix of the state of charge parameter.

[0249] After obtaining various battery state parameters, such as the measured mean and measurement covariance matrix of the state of charge (POC) parameter, through the aforementioned scheme, in this embodiment, the prior estimates of the POC parameter and the prior estimates of the POC parameter covariance matrix are updated based on the gain matrix corresponding to these information. This yields the updated POC parameter values, which typically include the posterior estimates of the POC parameter and the posterior estimates of the POC parameter covariance matrix. These posterior estimates will be used to calculate the prior estimates in the next round of updates. For example, based on the aforementioned content, the prior distribution value of the POC parameter in the next round of prior estimation can be determined using the posterior estimates of the POC parameter and the posterior estimates of the POC parameter covariance matrix.

[0250] Specifically, the calculation process for the posterior estimates of the charged state parameters and the posterior estimates of the covariance matrix of the charged state parameters included in the updated values ​​of the charged state parameters is as follows:

[0251] Calculate the Kalman filter gain matrix of the charged state parameter filtering system:

[0252] K k =P xy / P yy

[0253] The formula for calculating the posterior estimate of the charge state parameters is as follows:

[0254]

[0255] That is, the posterior estimate of the charged state parameters. For the prior estimate Based on this, add the actual voltage difference (i.e., the lowest single-section voltage U at time k) with the Kalman filter gain matrix as the coefficient. term With model predicted voltage Y k (the difference).

[0256] The formula for calculating the posterior estimate of the covariance matrix of the charged state parameters is as follows:

[0257]

[0258] In other words, the posterior estimate of the covariance matrix of the charged state parameters For the prior estimate Subtract the Kalman filter gain matrix K from the base k The third measurement covariance matrix P yy and the transpose of the Kalman filter gain matrix The result of matrix multiplication.

[0259] Of course, the above is only one feasible implementation scheme for updating the charge state parameters and obtaining the posterior distribution values ​​of the charge state parameters. Based on this, filtering innovation, that is, the difference between the observed value and the predicted value (U), can also be considered in the update process. term，k -Y k To perform multi-level judgments, for example, when |U term，k -Y k If |≤U1, it is determined to be an uncorrectable state. Let |U term，k -Y k |=0; when U1≤|U term，k -Y k If |≤U2, it is determined to be a normal correction state, |U term，k -Y k | unchanged; when U2≤|U term，k -Y k If |≤U3, it is determined to be a fault-prone correction state. Let |U term，k -Y k |=U3; when U3≤|U term，k -Y k If the process of posterior estimation is deemed to be flawed, exit the current process.

[0260] Of course, in addition to this, posterior estimates of the charged state parameters and the covariance matrix of the charged state parameters can also be introduced to determine whether the charged state parameter estimates need to be updated to the currently calculated posterior estimates. For example, in conjunction with the aforementioned scheme, if the posterior estimate of the covariance matrix of the charged state parameters is a symmetric, positive definite matrix, and / or, the posterior estimate of the covariance matrix of the charged state parameters is within a preset covariance matrix range, and / or, the posterior estimate of the charged state parameters is within a preset range of charged state parameters, then the posterior estimation result can be considered normal, and the charged state parameters can be updated normally. Otherwise, it can be considered that there is an abnormality in the posterior estimation process, and the previously calculated estimates, such as the prior estimates of the charged state parameters, are retained for subsequent calculations, such as the prior calculation process in the next round.

[0261] Of course, based on a similar processing procedure, in another embodiment, the posterior distribution estimation of the battery model parameters also involves posterior estimation of the parameters using the measurement mean and measurement covariance matrix. That is, the posterior estimation of the prior estimates of the battery model parameters based on the prior estimates of the state of charge parameters, to obtain updated values ​​of the battery model parameters in the battery state parameters, includes:

[0262] Based on the second measurement mean and the second gain matrix corresponding to the second measurement covariance matrix, the prior estimates of the battery model parameters and the prior estimates of the covariance matrix of the battery model parameters are updated to obtain updated values ​​of the battery model parameters. The updated values ​​of the battery model parameters include the posterior estimates of the battery model parameters and the posterior estimates of the covariance matrix of the battery model parameters.

[0263] The posterior calculation formula for the battery model parameters is basically the same as the posterior part of the aforementioned state-of-charge (POC) parameters. The corresponding input variables, namely the prior estimates of the POC parameters and the prior estimates of the POC parameter covariance matrix, need to be changed to the prior estimates of the battery model parameters. Prior estimates of the covariance matrix of the battery model parameters At this point, the formula for calculating the posterior estimates of the battery model parameters is:

[0264]

[0265] The parameters in the above formula have been explained in detail above, such as K. k The Kalman gain matrix used in the posterior estimation of the charge state parameters is adopted, and will not be repeated in the embodiments of this application.

[0266] Of course, in addition to the above, the posterior estimation of battery model parameters can also be based on the posterior estimates of the battery model parameters and the posterior estimates of the covariance matrix of the battery model parameters to determine whether the battery model parameter estimates need to be updated to the currently calculated posterior estimates. That is, if the posterior estimate of the covariance matrix of the battery model parameters is a symmetric, positive definite matrix, and / or the posterior estimate of the covariance matrix of the battery model parameters is within the preset covariance matrix range, and / or the posterior estimate of the battery model parameters is within the preset parameter range, the posterior estimation result can be considered normal, and the battery model parameters can be updated normally. Otherwise, the posterior estimation process can be considered abnormal, and the previously calculated estimates, such as the prior estimates of the battery model parameters, should be retained for subsequent calculations, such as the prior calculation process in the next round.

[0267] Alternatively, in another embodiment, the posterior distribution estimation of the capacity parameter also involves posterior estimation of the parameter using the measurement mean and measurement covariance matrix. That is, the prior estimate of the capacity parameter based on the prior estimate of the state of charge parameter is used to perform a posterior estimation of the prior estimate of the capacity parameter to obtain the updated value of the capacity parameter in the battery state parameters, including:

[0268] Based on the first measurement mean and the first gain matrix corresponding to the first measurement covariance matrix, the prior estimate of the capacity parameter and the prior estimate of the covariance matrix of the capacity parameter are updated to obtain the updated value of the capacity parameter. The updated value of the capacity parameter includes the posterior estimate of the capacity parameter and the posterior estimate of the covariance matrix of the capacity parameter.

[0269] Specifically, the posterior calculation formula for the capacity parameter is basically the same as the posterior part of the charge state parameter provided above. The corresponding input variables, namely the prior estimates of the charge state parameter and the prior estimates of the covariance matrix of the charge state parameter, need to be changed to the prior estimates of the capacity parameter. Prior estimates of the covariance matrix of the battery model parameters At this point, the formula for calculating the posterior estimate of the capacity parameter is:

[0270]

[0271] The parameters in the above formula have been explained in detail above, such as K. k The Kalman gain matrix used in the posterior estimation of the charge state parameters is adopted, and will not be repeated in the embodiments of this application.

[0272] Of course, in addition to the above, the posterior estimation of the capacity parameter can also be based on the posterior estimate of the capacity parameter and the posterior estimate of the covariance matrix of the capacity parameter to determine whether the capacity parameter estimate needs to be updated to the currently calculated posterior estimate. That is, if the posterior estimate of the covariance matrix of the capacity parameter is a symmetric, positive definite matrix, and / or the posterior estimate of the covariance matrix of the capacity parameter is within the preset covariance matrix range, and / or the posterior estimate of the capacity parameter is within the preset capacity range, the posterior estimation result can be considered normal, and the capacity parameter can be updated normally. Otherwise, the posterior estimation process can be considered abnormal, and the previously calculated estimate, such as the prior estimate of the capacity parameter, should be retained for subsequent calculations, such as the prior calculation process in the next round.

[0273] The aforementioned solution can fully complete the filtering and updating of various battery state parameters. Specifically, this application employs a square root unscented Kalman filter algorithm to achieve online identification and updating of model parameters, which improves computational accuracy and correction rate compared to the general recursive least squares algorithm. Compared to offline laboratory calibration of battery impedance parameters, online correction can eliminate cell inconsistencies and subsequent aging differences, improving the robustness of model parameters. Furthermore, compared to the extended Kalman filter algorithm, it eliminates Taylor expansion bias of second-order and higher nonlinear functions. Compared to conventional triggered correction, it increases the probability of correction in non-plateau regions, preventing sudden SOC jumps, and can also follow cell characteristics in real time. Compared to triggered correction strategies, it can improve correction rate and probability, and promptly address abnormal aging and inconsistency capacity deviation issues.

[0274] Building upon the aforementioned solutions, as a further feasible implementation of this application, this application also provides a method that employs a polarization level determination strategy to improve the correction probability and accuracy of the model in hybrid or regenerative braking modes, effectively eliminating operating conditions with high battery diffusion impedance. For example, in one embodiment, the method provided by this application further includes:

[0275] Based on the current information within a preset period, the charge and discharge state of the battery and the target polarity level corresponding to the charge and discharge state are determined; wherein, different polarity levels correspond to different correlation curves, and the correlation curves are the correlation curves between open circuit voltage and state of charge.

[0276] During the process of updating different battery state parameters based on different update frequencies, the battery state parameters are updated based on the target relationship curve corresponding to the target polarity level.

[0277] Specifically, in this embodiment, after the vehicle is connected to high voltage, the ampere-hour integral accumulation begins. If the discharge current is positive and the charging current is negative, the ampere-hour integral is reset to zero every fixed time T1. If the integral amount within T1 is ≥ C1, and the percentage of current ≥ I1 within the period is > A% & < B%, and the percentage of current < I2 is < C%, then it is determined to be first-order discharge polarization.

[0278] If the integral of C1≥T1 within the period is ≥C2, and the percentage of current ≥I1 within the period is >A% & <B, and the percentage of current <I2 is <C%, then it is determined to be second-order discharge polarization.

[0279] If the integral of C3 ≥ T1 ≥ C4, and the percentage of current ≤ I3 within the period is > A% & < B, and the percentage of current > I4 is < C%, then it is determined to be second-order charging polarization.

[0280] If the integral quantity within period T1 is ≤ C4, and the percentage of current ≤ I3 within the period is > A% & < B, and the percentage of current > I4 is < C%, then it is determined to be first-order charging polarization.

[0281] In this context, C1 and C2 are positive values, while C3 and C4 are negative values. Similarly, I1 and I2 are positive values, while I3 and I4 are negative values. After determining the battery's charge / discharge state and corresponding polarity level using different information, the OCV test schemes for different polarity levels differ. Therefore, in subsequent update processing, when it is necessary to use the SOC-OCV curve, such as obtaining the OCV lookup value through SOC, the target relationship curve corresponding to the target polarity level will be used to update the battery state parameters. By determining the polarization level, the cell polarization direction and polarization type in hybrid mode or regenerative braking mode can be accurately identified, improving the accuracy of the characteristic cell model.

[0282] If any of the above four conditions are met, the parameters can be updated using the method provided in this application in the next cycle. Otherwise, the battery diffusion impedance will be determined to be too large, and the SOC will be calculated using the ampere-hour integration method in the next cycle. At this time, the model internal resistance parameters and capacity remain unchanged.

[0283] Furthermore, considering that commonly used vehicle batteries, such as lithium iron phosphate cells, have a large plateau region in their static voltage, which can significantly impact battery pack state estimation, the solution provided in this application includes a step of determining whether to update the battery state parameters using the ratio of the change in open-circuit voltage to the change in state of charge over a period of time. In other words, the method further includes:

[0284] Determine the ratio of the change in open-circuit voltage to the change in state of charge within a preset period;

[0285] When the ratio is within a preset ratio range, the step of updating different battery state parameters based on different update frequencies is performed.

[0286] Specifically, if dOCV / dSOC is too small, it will affect the Kalman filter gain, causing the covariance matrix to diverge, making it impossible to correct the SOC, and consequently, the capacity. Conversely, if dOCV / dSOC is too large, the high degree of nonlinearity will lead to low accuracy in the filtering algorithm. Therefore, if dOCV / dSOC is within a suitable range, the collaborative estimation algorithm provided in this application is executed; otherwise, the ampere-hour integral method is used to calculate the SOC, in which case the model's internal resistance parameters and capacity remain unchanged.

[0287] Of course, the purpose of obtaining accurate battery state parameters provided in this application can be used to accurately assist vehicle control. That is, in a feasible implementation, predicting the battery state based on the updated battery state parameters includes:

[0288] The maximum allowable discharge current within a preset discharge cycle is determined based on the updated battery state parameters.

[0289] Based on the filtered values ​​of various battery state parameters obtained above, combined with the battery model, the maximum current value that the battery can provide to the load in the short term can be predicted, providing a basis for the vehicle's power control.

[0290] Current prediction only defines the first-level polarization of charging or discharging. A positive current indicates discharging, and a negative current indicates charging.

[0291] The zero-state response of the characteristic cell model is described as follows:

[0292]

[0293] Therefore, within time T2, the maximum allowable discharge current of the battery is:

[0294]

[0295] The discharge duration T2 can be adjusted according to the vehicle's power requirements and can be an integer multiple of 5 seconds. Different durations correspond to different acceleration and hill-climbing requirements. The current prediction period can be consistent with the SOC (State of Charge) parameter filtering period, using a third-order time scale. Alternatively, it can be consistent with the battery model parameter filtering period, using a second-order time scale. This application will not repeat the explanation.

[0296] Furthermore, considering that the charging process is mostly a unilateral current, the lack of rest or reverse discharge current during charging leads to excessive diffusion impedance in the lithium battery, thus causing the characteristic cell model to fail. Therefore, in this embodiment, a nonlinear connection algorithm for the charging end is also provided. That is, if it is determined that the charging end has been reached (SOC or voltage greater than a certain value) during the charging process, the SOC is calculated using the following formula:

[0297]

[0298] Among them, SOC t U represents the state of charge after time t during charging, and SOC0 represents the state of charge at the initial moment of charging. t Ut is the measured voltage at time t during charging, and U0 is the measured voltage at the initial moment of charging. end K1 and K2 are preset system gain coefficients, where K1 is the cutoff voltage.

[0299] The above methods can effectively address the issues of SOC jumps at the end of charging or SOC stagnation at a certain point for a long time. At the same time, the nonlinear connection method can also reduce the problem of long charging time in the plateau region of lithium iron phosphate cells.

[0300] This application updates different battery state parameters based on different update frequencies, that is, by updating battery state parameters through different time scales. This can better adapt to the update needs of different battery state parameters, so that the updated battery state parameters can be used more accurately for updating battery state parameters.

[0301] Figure 7 This is a block diagram illustrating an electronic device 700 according to an exemplary embodiment. Figure 7 As shown, the electronic device 700 may include a processor 701 and a memory 702. The electronic device 700 may also include one or more of a multimedia component 703, an input / output (I / O) component 704, and a communication component 705. In this embodiment, the electronic device 700 may be a device that integrates and implements the security event detection method provided in this embodiment.

[0302] The processor 701 controls the overall operation of the electronic device 700 to complete all or part of the steps in the aforementioned security event detection method. The memory 702 stores various types of data to support the operation of the electronic device 700. This data may include, for example, instructions for any application or method operating on the electronic device 700, and application-related data such as contact data, sent and received messages, pictures, audio, video, etc. The memory 702 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Read Only Memory (PROM), Read Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. The multimedia component 703 may include a screen and audio components. The screen may be, for example, a touchscreen, and the audio component is used to output and / or input audio signals. For example, the audio component may include a microphone for receiving external audio signals. The received audio signals may be further stored in memory 702 or transmitted via communication component 705. The audio component also includes at least one speaker for outputting audio signals. I / O component 704 provides an interface between processor 701 and other interface modules, such as a keyboard, mouse, buttons, etc. These buttons may be virtual or physical buttons. Communication component 705 is used for wired or wireless communication between the electronic device 700 and other devices. Wireless communication, such as Wi-Fi, Bluetooth, Near Field Communication (NFC), 2G, 3G, 4G, NB-IoT, eMTC, or other 5G technologies, or combinations thereof, is not limited here. Therefore, the corresponding communication component 705 may include: a Wi-Fi module, a Bluetooth module, an NFC module, etc.

[0303] In an exemplary embodiment, the electronic device 700 may be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field-programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to perform the battery state prediction method described above.

[0304] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the battery state prediction method provided in any of the above embodiments.

[0305] This application also provides a battery in which the battery state is determined by performing the steps of the battery state prediction method provided in any of the above embodiments.

[0306] This application also provides a vehicle equipped with the battery described above.

[0307] In one embodiment, the vehicle can be configured for fully or partially autonomous driving. For example, the vehicle can control itself while in autonomous driving mode, and can determine the current state of the vehicle and its surrounding environment through human intervention, determine the possible behaviors of at least one other vehicle in the surrounding environment, and determine the confidence level corresponding to the probability of that other vehicle performing a possible behavior, and control the vehicle based on the determined information. When the vehicle is in autonomous driving mode, it can be configured to operate without human interaction.

[0308] The vehicle may also include various subsystems, such as a driving system, sensor system control system, one or more peripheral devices, as well as power supply, computer system, and user interface. Optionally, the vehicle may include more or fewer subsystems, and each subsystem may include multiple components, such as multiple ECUs (electronic control units, i.e., vehicle computers) per subsystem.

[0309] In addition, each subsystem and component of the vehicle can be interconnected via wired or wireless means.

[0310] A propulsion system may include components that provide powered motion to the vehicle. In one embodiment, the propulsion system may include an engine, an energy source, a transmission, and wheels / tires. The engine may be an internal combustion engine, an electric motor, an air-compressed engine, or a combination of other types of engines, such as a hybrid engine consisting of a gasoline engine and an electric motor, or a hybrid engine consisting of an internal combustion engine and an air-compressed engine. The engine converts energy into mechanical energy.

[0311] Examples of energy sources include gasoline, diesel, other petroleum-based fuels, propane, other compressed gas-based fuels, ethanol, solar panels, batteries, and other sources of electricity. Energy sources can also power other systems in the vehicle.

[0312] A transmission system can transmit mechanical power from an engine to the wheels. The transmission system may include a gearbox, a differential, and a drive shaft. In one embodiment, the transmission system may also include other components, such as a clutch. The drive shaft may include one or more axles that can be coupled to one or more wheels.

[0313] A sensor system may include several sensors that sense information about the vehicle's surrounding environment. For example, a sensor system may include a positioning system (which could be GPS, BeiDou, or another positioning system), an inertial measurement unit (IMU), radar, a laser rangefinder, and cameras. The sensor system may also include sensors from the vehicle's internal systems being monitored (e.g., an in-vehicle air quality monitor, fuel gauge, oil temperature gauge, etc.). Sensor data from one or more of these sensors can be used to detect objects and their corresponding characteristics (position, shape, orientation, speed, etc.). This detection and identification is a critical function for the safe operation of autonomous vehicles.

[0314] A positioning system can be used to estimate a vehicle's geographical location. An IMU is used to sense changes in the vehicle's position and orientation based on inertial acceleration. In one embodiment, the IMU can be a combination of an accelerometer and a gyroscope.

[0315] Radar can use radio signals to sense objects in the vehicle's surrounding environment. In some embodiments, in addition to sensing objects, radar can also be used to sense the speed and / or direction of travel of objects.

[0316] A laser rangefinder can use lasers to sense objects in the environment in which a vehicle is located. In some embodiments, a laser rangefinder may include one or more laser sources, a laser scanner, one or more processing modules, and other system components.

[0317] The camera can be used to capture multiple images of the vehicle's surroundings. The camera can be a still camera or a video camera.

[0318] A control system controls the operation of a vehicle and its components. Control systems can include various elements, including steering systems, throttles, braking units, computer vision systems, route control systems, and obstacle avoidance systems.

[0319] The steering system is operable to adjust the vehicle's direction of travel. For example, in one embodiment, it can be a steering wheel system.

[0320] The throttle is used to control the engine's operating speed and, consequently, the vehicle's speed.

[0321] The braking unit is used to control the deceleration of the vehicle. The braking unit uses friction to slow down the wheels.

[0322] In other embodiments, the braking unit can convert the kinetic energy of the wheels into electrical current. The braking unit may also take other forms to slow down the wheel rotation speed, thereby controlling the vehicle speed.

[0323] Computer vision systems can be operated to process and analyze images captured by cameras to identify objects and / or features in the environment surrounding a vehicle. These objects and / or features may include traffic signals, road boundaries, and obstacles. Computer vision systems may use object recognition algorithms, structure from motion (SFM) algorithms, video tracking, and other computer vision techniques. In some embodiments, computer vision systems may be used to map the environment, track objects, estimate object velocities, and so on.

[0324] A route control system is used to determine the driving route of a vehicle. In some embodiments, the route control system may combine data from GPS and one or more predetermined maps to determine the driving route for the vehicle.

[0325] Obstacle avoidance systems are used to identify, assess, and avoid or otherwise traverse potential obstacles in the environment in which a vehicle is located.

[0326] In the description of this application, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include one or more features. In the description of this application, "multiple" means two or more, unless otherwise explicitly specified.

[0327] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0328] The embodiments, implementation methods, and related technical features of this application can be combined and substituted for each other without conflict.

[0329] The above are merely preferred embodiments of this application and are not intended to limit this application in any way. Any simple modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of this application without departing from the scope of the technical solution of this application shall still fall within the scope of the technical solution of this application.

Claims

1. A battery state prediction method characterized by, The method comprises: updating different battery state parameters based on different update frequencies to predict the battery state based on the updated battery state parameters, wherein the battery state parameters comprise at least two of a state of charge parameter, a battery model parameter and a capacity parameter.

2. The method of claim 1, wherein, The updating different battery state parameters based on different update frequencies comprises: updating different battery state parameters when accumulated time length and / or charge change amount in the battery charging and discharging process meet different conditions.

3. The method of claim 2, wherein, The updating different battery state parameters when accumulated time length and / or charge change amount in the battery charging and discharging process meet different conditions comprises: updating the capacity parameter in the battery state parameters when the accumulated time length in the battery charging and discharging process reaches a first preset time length and / or the charge change amount reaches a first preset change amount; or updating the battery model parameter in the battery state parameters when the accumulated time length in the battery charging and discharging process reaches a second preset time length and / or the charge change amount reaches a second preset change amount; or updating the state of charge parameter in the battery state parameters when the accumulated time length in the battery charging and discharging process reaches a third preset time length and / or the charge change amount reaches a third preset change amount.

4. The method of claim 3, wherein, The first preset time length is greater than the second preset time length, the second preset time length is greater than the third preset time length, the first preset change amount is greater than the second preset change amount, and the second preset change amount is greater than the third preset change amount.

5. The method of claim 1, wherein, The updating different battery state parameters comprises: updating the capacity parameter in the battery state parameters based on the capacity parameter before updating and a priori estimated value of the state of charge parameter; or updating the battery model parameter in the battery state parameters based on the battery model parameter before updating and a priori estimated value of the state of charge parameter; or updating the state of charge parameter in the battery state parameters based on the state of charge parameter before updating, a priori estimated value of the capacity parameter and a priori estimated value of the battery model parameter.

6. The method of claim 5, wherein, The updating the capacity parameter in the battery state parameters based on the capacity parameter before updating and a priori estimated value of the state of charge parameter comprises: performing a priori estimation on the capacity parameter before updating by a preset filtering algorithm to obtain the a priori estimated value of the capacity parameter; performing a posteriori estimation on the a priori estimated value of the capacity parameter based on the a priori estimated value of the state of charge parameter to obtain the updated value of the capacity parameter in the battery state parameters.

7. The method of claim 6, wherein, The performing a priori estimation on the capacity parameter before updating by a preset filtering algorithm to obtain the a priori estimated value of the capacity parameter comprises: obtaining a plurality of a priori distribution values of the capacity parameter based on a covariance matrix of the capacity parameter before updating; obtaining the a priori estimated value of the capacity parameter and the a priori estimated value of the covariance matrix of the capacity parameter based on the plurality of a priori distribution values of the capacity parameter. The prior estimation value of the capacity parameter is updated in a case that the prior estimation value of the covariance matrix of the capacity parameter and / or the prior estimation value of the capacity parameter satisfies a preset condition.

8. The method of claim 7, wherein, The prior estimation value of the capacity parameter is obtained based on the plurality of prior distribution values of the capacity parameter, and the prior estimation value of the covariance matrix of the capacity parameter comprises: The prior estimation value of the capacity parameter is obtained by weighting a result obtained by processing the prior distribution value of the capacity parameter through a capacity filtering state equation based on a mean weight corresponding to the prior distribution value of the capacity parameter; The prior estimation value of the covariance matrix of the capacity parameter is obtained by weighting a difference between the prior distribution value of the capacity parameter and the prior estimation value of the capacity parameter based on a covariance weight corresponding to the prior distribution value of the capacity parameter.

9. The method of claim 8, wherein, The prior estimation value of the capacity parameter is updated in a case that the prior estimation value of the covariance matrix of the capacity parameter and / or the prior estimation value of the capacity parameter satisfies a preset condition. The prior estimation value of the capacity parameter is updated in a case that the prior estimation value of the covariance matrix of the capacity parameter is a negative definite matrix, the prior estimation value of the covariance matrix of the capacity parameter is a positive definite matrix, and / or the prior estimation value of the covariance matrix of the capacity parameter is within a preset covariance boundary range and / or the prior estimation value of the capacity parameter is within a preset capacity boundary range.

10. The method of claim 6, wherein, The updated value of the capacity parameter is obtained by performing a posterior estimation on the prior estimation value of the capacity parameter based on the prior estimation value of the state of charge parameter. A plurality of posterior distribution values of the capacity parameter are obtained based on the prior estimation value of the capacity parameter and the prior estimation value of the covariance matrix of the capacity parameter. A first measurement mean value and a first measurement covariance matrix are obtained by processing the plurality of posterior distribution values of the capacity parameter and the prior estimation value of the state of charge parameter through a capacity filtering measurement equation. The prior estimation value of the capacity parameter and the prior estimation value of the covariance matrix of the capacity parameter are updated based on a first gain matrix corresponding to the first measurement mean value and the first measurement covariance matrix to obtain an updated value of the capacity parameter, wherein the updated value of the capacity parameter comprises a posterior estimation value of the capacity parameter and a posterior estimation value of the covariance matrix of the capacity parameter.

11. The method of claim 5, wherein, The battery model parameter is updated based on the prior estimation value of the battery model parameter before updating and the prior estimation value of the state of charge parameter. The prior estimation value of the battery model parameter is obtained by performing a prior estimation on the battery model parameter before updating through a preset filtering algorithm. The updated value of the battery model parameter is obtained by performing a posterior estimation on the prior estimation value of the battery model parameter based on the prior estimation value of the state of charge parameter.

12. The method of claim 11, wherein, The prior estimation value of the battery model parameter is obtained by performing a prior estimation on the battery model parameter before updating based on a covariance matrix of the battery model parameter before updating. The plurality of prior distribution values of the battery model parameter are obtained based on the covariance matrix of the battery model parameter before updating. obtaining a prior estimation value of the battery model parameter and a prior estimation value of a covariance matrix of the battery model parameter based on a plurality of prior distribution values of the battery model parameter; updating the prior estimation value of the battery model parameter when the prior estimation value of the covariance matrix of the battery model parameter and / or the prior estimation value of the battery model parameter satisfies a preset condition.

13. The method of claim 12, wherein, The method comprises the following steps: weighting a result obtained by processing the prior distribution value of the battery model parameter through a battery model filtering state equation based on a mean weight corresponding to the prior distribution value of each battery model parameter, to obtain the prior estimation value of the battery model parameter; weighting a difference between the prior distribution value of the battery model parameter and the prior estimation value of the battery model parameter based on a covariance weight corresponding to the prior distribution value of each battery model parameter, to obtain the prior estimation value of the covariance matrix of the battery model parameter.

14. The method of claim 12, wherein, The method comprises the following steps: updating the prior estimation value of the battery model parameter when the prior estimation value of the covariance matrix of the battery model parameter is a negative definite matrix, a positive definite matrix, and / or the prior estimation value of the covariance matrix of the battery model parameter is within a preset covariance boundary range and / or the prior estimation value of the battery model parameter is within a preset model parameter boundary range.

15. The method of claim 11, wherein, The method comprises the following steps: obtaining a plurality of posterior distribution values of the battery model parameter based on the prior estimation value of the battery model parameter and the prior estimation value of the covariance matrix of the battery model parameter; processing the plurality of posterior distribution values of the battery model parameter and the prior estimation value of the state of charge parameter through a battery model filtering measurement equation to obtain a second measurement mean value and a second measurement covariance matrix; updating the prior estimation value of the battery model parameter and the prior estimation value of the covariance matrix of the battery model parameter based on a second gain matrix corresponding to the second measurement mean value and the second measurement covariance matrix to obtain an updated value of the battery model parameter, wherein the updated value of the battery model parameter comprises a posterior estimation value of the battery model parameter and a posterior estimation value of the covariance matrix of the battery model parameter.

16. The method of claim 5, wherein, The method comprises the following steps: obtaining the prior estimation value of the state of charge parameter by performing prior estimation on the state of charge parameter before updating through a preset filtering algorithm; obtaining the prior estimation value of the state of charge parameter by performing prior estimation on the state of charge parameter before updating through a preset filtering algorithm; The prior estimation value of the state-of-charge parameter is updated based on the prior estimation value of the capacity parameter and the prior estimation value of the battery model parameter.

17. The method of claim 16, wherein, The prior estimation value of the state-of-charge parameter is obtained by performing prior estimation on the state-of-charge parameter before updating, including: A plurality of prior distribution values of the state-of-charge parameter are obtained based on a covariance matrix of the state-of-charge parameter before updating; The prior estimation value of the state-of-charge parameter and the prior estimation value of the covariance matrix of the state-of-charge parameter are obtained based on the plurality of prior distribution values of the state-of-charge parameter; The prior estimation value of the state-of-charge parameter is updated when the prior estimation value of the covariance matrix of the state-of-charge parameter and / or the prior estimation value of the state-of-charge parameter satisfies a preset condition.

18. The method of claim 17, wherein, The prior estimation value of the state-of-charge parameter and the prior estimation value of the covariance matrix of the state-of-charge parameter are obtained based on the plurality of prior distribution values of the state-of-charge parameter, including: The result obtained by processing the prior distribution value of the state-of-charge parameter through a state-of-charge filtering state equation is weighted based on a mean weight corresponding to each prior distribution value of the state-of-charge parameter, to obtain the prior estimation value of the state-of-charge parameter; The difference between the prior distribution value of the state-of-charge parameter and the prior estimation value of the state-of-charge parameter is weighted based on a covariance weight corresponding to each prior distribution value of the state-of-charge parameter, to obtain the prior estimation value of the covariance matrix of the state-of-charge parameter.

19. The method of claim 17, wherein, The prior estimation value of the state-of-charge parameter is updated when the prior estimation value of the covariance matrix of the state-of-charge parameter and / or the prior estimation value of the state-of-charge parameter satisfies a preset condition, including: The prior estimation value of the state-of-charge parameter is updated when the prior estimation value of the covariance matrix of the state-of-charge parameter is a negative definite matrix, the prior estimation value of the covariance matrix of the state-of-charge parameter is within a preset covariance boundary range, and / or the prior estimation value of the state-of-charge parameter is within a preset state-of-charge boundary range.

20. The method of claim 16, wherein, The prior estimation value of the state-of-charge parameter is updated based on the prior estimation value of the capacity parameter and the prior estimation value of the battery model parameter. A plurality of posterior distribution values of the state-of-charge parameter are obtained based on the prior estimation value of the state-of-charge parameter and the prior estimation value of the covariance matrix of the state-of-charge parameter; A third measurement mean and a third measurement covariance matrix are obtained by processing the plurality of posterior distribution values of the state-of-charge parameter, the prior estimation value of the capacity parameter and the prior estimation value of the battery model parameter through a state-of-charge filtering measurement equation; updating the prior estimation of the state of charge parameter and the prior estimation of the covariance matrix of the state of charge parameter based on the third measurement mean value and the third gain matrix corresponding to the third measurement covariance matrix, to obtain an updated value of the state of charge parameter, the updated value of the state of charge parameter including a posteriori estimation of the state of charge parameter and a posteriori estimation of the covariance matrix of the state of charge parameter.

21. The method of claim 1, wherein, The method further comprises: determining a charge-discharge state of the battery and a target polarity level corresponding to the charge-discharge state based on current information in a preset period; wherein different polarity levels correspond to different correlation curves, and the correlation curve is a correlation curve between open circuit voltage and state of charge; and in the process of updating different battery state parameters based on different update frequencies respectively, updating the battery state parameters based on a target relationship curve corresponding to the target polarity level.

22. The method of claim 1, wherein, The method further comprises: determining a ratio of a change amount of the open circuit voltage to a change amount of the state of charge in a preset period; when the ratio is within a preset ratio range, performing the step of updating different battery state parameters based on different update frequencies respectively.

23. The method of claim 22, wherein, The method further comprises: when the ratio is outside the preset ratio range, determining a predicted battery state based on current integration.

24. The method of claim 1, wherein, The method further comprises: determining a maximum allowable discharge current in a preset discharge period based on the updated battery state parameter.

25. The method of any one of claims 1-24, wherein, The method further comprises: in a charging end process, predicting the state of charge parameter of the battery by the following formula: wherein SOC t is the state of charge after charging, SOC0 is the state of charge at the beginning of charging, U t is the measured voltage at the time t of charging, U0 is the measured voltage at the beginning of charging, U end is the cut-off voltage, and K1 and K2 are preset system gain coefficients.

26. An electronic device, comprising: a processor; a memory for storing processor executable instructions; wherein the processor is configured to perform the steps of the safety event detection method of any one of claims 1-25.

27. A computer-readable storage medium, characterized in that, a computer program is stored thereon, and the computer program is loaded by a processor to perform the steps of the safety event detection method of any one of claims 1-25.

28. A battery, characterized by The battery determines the battery state of the battery by performing the steps of the battery state prediction method of any one of claims 1-25.

29. A vehicle characterized by a battery as claimed in claim 28. a battery as claimed in claim 28.

Images