Battery parameter online identification method, device and equipment

Abstract

The application provides a battery parameter online identification method, device and equipment, and relates to the technical field of battery management system, and comprises the following steps: predicting future working condition data of a battery BMS system based on historical working condition data of the battery BMS system; determining a target parameter identification mode based on the future working condition data of the battery BMS system; in the case that the future working condition data triggers switching from a current parameter identification mode to the target parameter identification mode, updating parameters of an online parameter identification algorithm by using reference parameters corresponding to the target parameter identification mode, wherein the parameters of the online parameter identification algorithm at least include a forgetting factor; and identifying to-be-identified parameters of the battery BMS system based on the updated forgetting factor and a target discrete equivalent circuit model corresponding to the battery BMS system by using the online parameter identification algorithm to obtain a parameter identification result. The application can significantly improve the accuracy of parameter identification of the battery BMS system and meet the actual application requirements.

Description

Technical Field

[0001] This invention relates to the field of battery management system technology, and in particular to a method, apparatus and equipment for online identification of battery parameters. Background Technology

[0002] As a widely used energy storage device, the performance evaluation and state monitoring of lithium-ion batteries rely on accurate battery parameter identification. In applications such as electric vehicles and energy storage systems, the precise acquisition of battery parameters is crucial for the optimized control of the battery management system (BMS).

[0003] In existing technologies, two traditional methods are mainly used for lithium-ion battery parameter identification:

[0004] The first method involves conducting standard charge-discharge experiments such as HPPC (Hybrid Pulse Power Characterization) on the battery using battery charging and discharging equipment, and then identifying parameters offline based on the experimental data. This method initially provides five fixed battery parameters and cannot achieve real-time parameter updates, failing to adapt to the time-varying characteristics of battery parameters. Furthermore, this method requires complete charge-discharge data, making it difficult to dynamically adjust battery parameters based on factors such as battery state and environmental conditions in practical applications, and thus cannot fully describe the changes in battery state.

[0005] The second method is based on the Extended Kalman Filter (EKF) technique, which jointly estimates battery parameters as state variables. While this method enables online parameter identification, it requires establishing a complex nonlinear state-space model and calculating the Jacobian matrix, significantly increasing computational complexity. This not only places high demands on the computational capabilities of the embedded system but also necessitates accurate acquisition of the statistical characteristics of process and observation noise. In practical applications, this method suffers from slow parameter convergence and susceptibility to initial value influence.

[0006] Both of the above methods have obvious limitations: they cannot fully describe the parameter changes of the battery throughout the entire process and under all conditions, and their high computational complexity makes it difficult to meet real-time requirements. Especially under complex operating conditions such as battery aging and temperature changes, the parameter identification accuracy and adaptability of existing methods often fail to meet the needs of practical applications. Summary of the Invention

[0007] In view of this, the purpose of the present invention is to provide a method, apparatus and device for online identification of battery parameters, which can significantly improve the accuracy of battery BMS system parameter identification and meet the needs of practical applications.

[0008] In a first aspect, the present invention provides a method for online identification of battery parameters, comprising:

[0009] Based on historical operating data of the battery BMS system, predict future operating data of the battery BMS system;

[0010] Based on future operating condition data from the battery BMS system, determine the target parameter identification mode;

[0011] When switching from the current parameter identification mode to the target parameter identification mode is triggered based on future working condition data, the parameters of the online parameter identification algorithm are updated using the benchmark parameters corresponding to the target parameter identification mode. The parameters of the online parameter identification algorithm include at least the forgetting factor.

[0012] The online parameter identification algorithm, based on the updated forgetting factor and combined with the target discretized equivalent circuit model corresponding to the battery BMS system, identifies the parameters to be identified in the battery BMS system and obtains the parameter identification results.

[0013] In one implementation, predicting future operating conditions of the battery BMS system based on historical operating condition data includes:

[0014] Based on a pre-trained operating condition prediction model, and using historical operating condition data of the battery BMS system, the future operating condition data of the battery BMS system is predicted.

[0015] In one implementation, future operating condition data includes current data, voltage data, and battery resting time; based on the future operating condition data of the battery BMS system, a target parameter identification mode is determined, including:

[0016] Match one or more of the current data, voltage data, and battery rest time with the trigger conditions of multiple pre-configured parameter identification modes;

[0017] The matched parameter identification pattern is determined as the target parameter identification pattern.

[0018] In one implementation, the parameter identification modes include a static mode, a quasi-steady-state mode, a transient mode, and an invalid data mode; matching one or more of current data, voltage data, and battery resting time with the trigger conditions of a plurality of pre-configured parameter identification modes, including:

[0019] If the current data is a specified value and the battery resting time exceeds a preset time threshold, it is determined that the triggering conditions match the resting mode.

[0020] If the rate of change of the current data is within the preset fluctuation range, determine the matching with the triggering conditions of the quasi-steady-state mode;

[0021] If the rate of change of the current data exceeds the preset fluctuation range, determine the matching with the triggering conditions of the transient mode;

[0022] In the event of abnormal current data and / or abnormal voltage data, determine whether the triggering conditions match those of the standby mode.

[0023] In one implementation, the parameters of the online parameter identification algorithm are updated using the baseline parameters corresponding to the target parameter identification pattern, including:

[0024] When the target parameter identification mode is a quasi-steady-state mode, the forgetting factor in the online parameter identification algorithm is set to the first value, and the update frequency in the online parameter identification algorithm is set to the first update frequency.

[0025] When the target parameter identification mode is transient mode, the forgetting factor in the online parameter identification algorithm is set to a second value, and the update frequency in the online parameter identification algorithm is set to a second update frequency; wherein, the second value is lower than the first value, and the second update frequency is higher than the first update frequency;

[0026] When the target parameter identification mode is the backup mode, the forgetting factor and update frequency in the online parameter identification algorithm remain unchanged.

[0027] In one implementation, the method further includes:

[0028] When the target parameter identification mode is set to static mode, the identification of the parameters to be identified in the battery BMS system is stopped.

[0029] In one implementation, an online parameter identification algorithm, based on an updated forgetting factor and combined with the target discretized equivalent circuit model corresponding to the battery BMS system, identifies the parameters to be identified in the battery BMS system to obtain parameter identification results, including:

[0030] A data vector is constructed based on the target discretized equivalent circuit model corresponding to the battery BMS system. The parameters in the data vector are state variables associated with the parameters to be identified in the battery BMS system. The state variables are determined based on the current operating condition data.

[0031] The updated online parameter identification algorithm performs the following operations during the current recursive process:

[0032] Based on the error covariance matrix from the previous recursive process, the updated forgetting factor and data vector from the online parameter identification algorithm, the gain matrix in the current recursive process is determined.

[0033] Based on the parameters to be identified in the previous recursive process, the gain matrix and data vector in the current recursive process, the parameters to be identified in the current recursive process are determined, so as to obtain the parameter identification results in the current recursive process.

[0034] The error covariance matrix in the current recursion process is determined based on the error covariance matrix from the previous recursion process, the gain matrix, the forgetting factor, and the data vector.

[0035] In one implementation, the method further includes:

[0036] Construct an initial equivalent circuit model corresponding to the battery BMS system, and discretize the initial equivalent circuit model to obtain an initial discretized equivalent circuit model.

[0037] The initial discretized equivalent circuit model is subjected to Laplace transform to obtain the frequency domain expression of the transfer function.

[0038] The frequency domain expression transfer function is subjected to bilinear transformation, and the target discretized equivalent circuit model is obtained by combining the bilinearly transformed frequency domain expression transfer function with the initial equivalent circuit model.

[0039] Secondly, the present invention also provides an online battery parameter identification device, comprising:

[0040] The operating condition prediction module is used to predict the future operating condition data of the battery BMS system based on the historical operating condition data of the battery BMS system.

[0041] The pattern determination module is used to determine the target parameter identification pattern based on future operating condition data of the battery BMS system.

[0042] The mode switching and parameter update module is used to update the parameters of the online parameter identification algorithm by using the benchmark parameters corresponding to the target parameter identification mode when the current parameter identification mode is switched to the target parameter identification mode based on future working condition data. The parameters of the online parameter identification algorithm include at least the forgetting factor.

[0043] The parameter identification module is used to identify the parameters to be identified in the battery BMS system by using an online parameter identification algorithm, based on the updated forgetting factor and combined with the target discretized equivalent circuit model corresponding to the battery BMS system, to obtain the parameter identification results.

[0044] Thirdly, the present invention also provides an electronic device including a processor and a memory, the memory storing computer-executable instructions executable by the processor, the processor executing the computer-executable instructions to implement any of the methods provided in the first aspect.

[0045] This invention provides a method, apparatus, and device for online battery parameter identification. First, based on historical operating condition data of the battery BMS system, future operating condition data of the battery BMS system is predicted. Then, based on the future operating condition data of the battery BMS system, a target parameter identification mode is determined. When switching from the current parameter identification mode to the target parameter identification mode is triggered by the future operating condition data, the parameters of the online parameter identification algorithm are updated using the benchmark parameters corresponding to the target parameter identification mode. The parameters of the online parameter identification algorithm include at least a forgetting factor. Finally, using the online parameter identification algorithm, based on the updated forgetting factor and combined with the target discretized equivalent circuit model corresponding to the battery BMS system, the parameters to be identified in the battery BMS system are identified to obtain the parameter identification result. The above method predicts future operating conditions based on historical operating data of the battery BMS system, determines the target parameter identification mode corresponding to the future operating conditions, updates the parameters of the online parameter identification algorithm using the benchmark parameters corresponding to the target parameter identification model, and identifies the parameters to be identified by combining the target discretized equivalent circuit model of the battery BMS system. This effectively improves the accuracy of parameter identification and provides accurate and effective battery parameters for subsequent battery SOX estimation in the battery BMS system.

[0046] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention are realized and obtained in accordance with the structures particularly pointed out in the description, claims and drawings.

[0047] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0048] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0049] Figure 1 A flowchart illustrating an online battery parameter identification method provided in an embodiment of the present invention;

[0050] Figure 2 A schematic diagram of a second-order RC equivalent circuit model provided in an embodiment of the present invention;

[0051] Figure 3 A flowchart of an online parameter identification process for FFRLS provided in an embodiment of the present invention;

[0052] Figure 4 A flowchart illustrating another online battery parameter identification method provided in an embodiment of the present invention;

[0053] Figure 5 This is a schematic diagram of the structure of an online battery parameter identification device provided in an embodiment of the present invention;

[0054] Figure 6 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0055] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0056] Currently, existing battery parameter identification methods cannot fully describe the parameter situation of the battery throughout the entire process and under all conditions, and the computational complexity is high. The accuracy and adaptability of parameter identification are often difficult to meet the needs of practical applications. Based on this, the present invention provides an online battery parameter identification method, device and equipment, which can significantly improve the accuracy of battery BMS system parameter identification and meet the needs of practical applications.

[0057] To facilitate understanding of this embodiment, a detailed description of the online battery parameter identification method disclosed in this embodiment of the invention will be provided first. (See [link to relevant documentation]). Figure 1 The diagram shows a flowchart of an online battery parameter identification method, which mainly includes the following steps S102 to S108:

[0058] Step S102: Based on the historical operating data of the battery BMS system, predict the future operating data of the battery BMS system.

[0059] Historical operating condition data and future operating condition data both include current data, voltage data, and battery resting time. In one implementation, a pre-trained operating condition prediction model can be used to predict the operating condition data (i.e., future operating condition data) of the battery BMS system in the future based on historical operating condition data from multiple historical time periods.

[0060] Step S104: Determine the target parameter identification mode based on the future operating condition data of the battery BMS system.

[0061] The parameter identification mode is used to adaptively adjust key parameters (such as forgetting factor, update frequency, etc.) of online parameter calculation for different future operating conditions. The parameter identification mode can include a static mode, a quasi-steady-state mode, a transient mode, and a standby mode. The static mode is when the current data is 0 and the battery has been idle for more than a preset time threshold. The quasi-steady-state mode is when the rate of change of the current data is within a preset fluctuation range. The transient mode is when the rate of change of the current data exceeds the preset fluctuation range. The standby mode is when there are anomalies in the current data and / or voltage data. The current parameter identification mode is the parameter identification mode at the current moment, and the target parameter identification mode is the parameter identification mode to be switched to.

[0062] In one example, future operating condition data can be matched with the triggering conditions of each parameter identification mode to determine the parameter identification mode that matches the future operating condition data, and this mode can be used as the target parameter identification mode.

[0063] Step S106: When switching from the current parameter identification mode to the target parameter identification mode is triggered based on future working condition data, the parameters of the online parameter identification algorithm are updated using the benchmark parameters corresponding to the target parameter identification mode. The parameters of the online parameter identification algorithm include at least a forgetting factor.

[0064] In one example, if the parameter identification mode matched by future operating data is inconsistent with the current parameter identification mode, a mode switch can be performed, that is, the current parameter identification mode can be switched to the target parameter identification mode. Technicians pre-configure corresponding key parameters for each parameter identification mode. When it is determined to switch to the target parameter identification model, the key parameters corresponding to the mode can be used to replace the current parameters of the online parameter identification algorithm. Optionally, the online parameter identification algorithm can adopt FFRLS (Forgetting Factor Recursive Least Squares).

[0065] Step S108: Using an online parameter identification algorithm, based on the updated forgetting factor and combined with the target discretized equivalent circuit model corresponding to the battery BMS system, the parameters to be identified in the battery BMS system are identified to obtain the parameter identification results.

[0066] The target discretized equivalent circuit model is used to describe the dynamic characteristics of the battery management system (BMS). In one example, a second-order RC equivalent circuit model corresponding to the battery BMS system is first constructed. Then, this second-order RC equivalent circuit model is discretized, subjected to Laplace transform, and bilinear transform to obtain the target discretized equivalent circuit model. Subsequently, using the FFRLS algorithm and its updated parameters (such as the forgetting factor), combined with the above target discretized equivalent circuit model, the parameters to be identified in the battery BMS system are identified to obtain the parameter identification results. The parameters to be identified may include capacitors, resistors, etc. in the battery BMS system, and the parameter identification results are the specific values ​​corresponding to the above-mentioned parameters.

[0067] The online battery parameter identification method provided in this invention predicts future operating conditions based on historical operating condition data of the battery BMS system, determines the target parameter identification mode corresponding to the future operating condition data, updates the parameters of the online parameter identification algorithm using the benchmark parameters corresponding to the target parameter identification model, and identifies the parameters to be identified by combining the target discretized equivalent circuit model of the battery BMS system. This effectively improves the accuracy of parameter identification and provides accurate and effective battery parameters for subsequent battery SOX estimation in the battery BMS system.

[0068] For ease of understanding, this embodiment of the invention provides a specific implementation method for an online battery parameter identification method.

[0069] (I) This invention provides a specific implementation method for constructing a target discretized equivalent circuit model corresponding to a battery BMS system, including:

[0070] Step 1.1: Construct the initial equivalent circuit model corresponding to the battery BMS system, and discretize the state space equation of the initial equivalent circuit model to obtain the initial discretized equivalent circuit model.

[0071] In one embodiment, the present invention uses a second-order RC equivalent circuit model as the initial equivalent circuit model of the battery BMS system, such as... Figure 2 The diagram shows a schematic of a second-order RC equivalent circuit model, which is based on the open-circuit voltage. ,resistance It consists of three parts: a resistor and two RC circuits connected in series. The ohmic internal resistance of the battery is expressed as , where , These are electrochemical polarization resistance and capacitance, respectively. , These are the concentration gradient polarization resistor and capacitor, respectively; a parallel RC circuit is used to simulate the polarization effect of the battery, and , The loop represents the impedance encountered by lithium ions during transport between two electrodes, simulating the short-term change in terminal voltage when encountering a sudden current surge. , The loop represents the process by which the terminal voltage changes slowly when it encounters a sudden current change. It is the output voltage of the battery; , These represent two RC rings (including) , Ring road and , The terminal voltage of the loop. This model can describe the dynamic characteristics of a real battery relatively accurately, and its state-space equation is expressed as:

[0072] ;

[0073] in, This indicates that the battery management system is at time [time missing]. The current data is used. The state-space equations are discretized to obtain the expression for the initial discretized equivalent circuit model:

[0074] ;

[0075] ;

[0076] in, = , = These are the time constants of the two RC loops, respectively. The sampling time interval, Sampling time, For battery capacity, C represents the remaining battery power. This is current data.

[0077] Step 1.2: Perform Laplace transform on the initial discretized equivalent circuit model to obtain the frequency domain expression transfer function.

[0078] In one implementation, the spatial state equation of the initial discretized equivalent circuit model (i.e., the discretized second-order RC battery model) is converted into a frequency domain expression through a Laplace transform, and its transfer function is obtained as follows:

[0079] ;

[0080] in, For the complex frequency variable in the Laplace transform, , , , , These are the parameters to be identified.

[0081] Step 1.3: Perform bilinear transformation on the frequency domain expression transfer function, and combine the bilinearly transformed frequency domain expression transfer function with the initial equivalent circuit model to obtain the target discretized equivalent circuit model.

[0082] In one implementation, the above transfer function is subjected to a bilinear transformation to obtain:

[0083] ;

[0084] in, For the new parameters to be identified, This is a discretization operator. Further parameters to be identified. For ease of calculation, the intermediate parameters are defined as follows:

[0085] ;

[0086] ;

[0087] make Combining this with the above equation yields the discretization formula for the second-order RC equivalent circuit model (i.e., the target discretized equivalent circuit):

[0088] ;

[0089] make , The system output equation is obtained as follows:

[0090] ;

[0091] in, For a data vector, the elements in the data vector , , , , The state variables (also known as the coefficients of the parameters to be identified) are associated with the parameters to be identified in the battery BMS system. This is the output vector of the identification result.

[0092] (II) This invention provides a specific implementation method for predicting the future operating condition data of a battery BMS system based on historical operating condition data. It uses a pre-trained operating condition prediction model to predict the future operating condition data of the battery BMS system based on historical operating condition data. Specifically, an LSTM (Long Short-Term Memory) model can be used. The specific implementation process is as follows:

[0093] The trained neural network model (i.e., the operating condition prediction model) predicts the voltage and current change trends of the battery BMS system over a period of time. Based on the predicted future operating condition data, it helps to judge the current battery operating condition and predict the future battery operating state. The predicted future operating condition data is further used as feedback input to dynamically trigger the mode switching mechanism and adjust the parameter configuration of the recursive least squares (FFRLS) algorithm with forgetting factor, thereby improving the system's ability to adapt to time-varying battery characteristics.

[0094] The input to the operating condition prediction model includes historical voltage data, historical current data, and battery resting time, such as current over 5-10 historical periods. ) sequence and voltage ( The output of the operating condition prediction model is the trend of battery voltage and current changes over a future period. The expression for the operating condition prediction model is shown below:

[0095] ;

[0096] ;

[0097] in, This represents the hidden state of the LSTM network. It is the current input data (i.e., historical operating condition data). It is the output data of the LSTM network (i.e., future operating condition data). , , It is the weight matrix of the LSTM network. , It is a bias term. It is an activation function.

[0098] (III) This invention provides a specific implementation method for determining target parameter identification patterns based on future operating condition data from a battery management system (BMS), including:

[0099] Step 3.1: Match one or more of the current data, voltage data, and battery resting time with the trigger conditions of multiple pre-configured parameter identification modes. This embodiment of the invention monitors the battery's current data, battery data, and battery resting time in real time, classifying the battery's operating conditions into the following modes: resting mode, quasi-steady-state mode, transient mode, and standby mode. These modes are collectively referred to as parameter identification modes.

[0100] In practice:

[0101] (1) When the current data is a specified value (such as 0) and the battery rest time exceeds a preset time threshold, determine that it matches the triggering condition of the rest mode. For example, when the current data is 0 and the battery rest time exceeds the set threshold of 2 minutes, determine to enter the rest mode.

[0102] (2) When the rate of change of the current data is within the preset fluctuation range, determine whether it matches the triggering condition of the quasi-steady-state mode. For example, when the current data is continuous and changes within the fluctuation range of 2A, determine whether to enter the quasi-steady-state mode.

[0103] (3) If the rate of change of the current data exceeds the preset fluctuation range, determine that it matches the triggering condition of the transient mode. For example, if the current data persists and the rate of change exceeds 2A, determine that the transient mode is entered.

[0104] (4) In the event of abnormal current data and / or abnormal voltage data, determine whether the system matches the triggering conditions of the standby mode. For example, if the voltage and current are both 0 during system operation, then determine whether to enter standby mode.

[0105] Step 3.2: Determine the matched parameter identification mode as the target parameter identification mode. For example, assuming the current parameter identification mode is a quasi-steady-state mode, while the mode identified by the operating condition data is a transient mode, the mode will be switched from the quasi-steady-state mode to the transient mode.

[0106] (iv) When the current parameter identification mode and the target parameter identification mode are inconsistent, a switch from the current parameter identification mode to the target parameter identification mode will be triggered. Based on this, this embodiment of the invention provides an implementation method for updating the parameters of the online parameter identification algorithm using the benchmark parameters corresponding to the target parameter identification mode, including:

[0107] Step 4.1: When the target parameter identification mode is a quasi-steady-state mode, set the forgetting factor in the online parameter identification algorithm to the first value, and set the update frequency in the online parameter identification algorithm to the first update frequency. For example, set the forgetting factor... Furthermore, a lower update frequency is set, such as updating the parameters to be identified once every 10 sampling periods, to increase the system's weight on historical data and thus increase the stability of the identification results.

[0108] Step 4.2: When the target parameter identification mode is transient mode, the forgetting factor in the online parameter identification algorithm is set to a second value, and the update frequency in the online parameter identification algorithm is set to a second update frequency; wherein, the second value is lower than the first value, and the second update frequency is higher than the first update frequency. The transient mode focuses on the accurate identification of dynamic parameters to be identified, hence the setting of the forgetting factor. 85, and a higher update frequency, such as updating the parameters to be identified in each sampling period, improves the response speed of the algorithm and increases the accuracy of the model.

[0109] Step 4.3: When the target parameter identification mode is in standby mode, keep the forgetting factor and update frequency in the online parameter identification algorithm unchanged. When the sensor malfunctions (e.g., voltage or current is 0) and the predicted result differs significantly from the measured data, it can be determined that the sensor has failed. In this case, the predicted result is used as the standard, the system switches to standby mode, parameter updates are paused, and the parameter values ​​from the previous moment are used, thereby avoiding contamination of the parameter identification results by erroneous data.

[0110] Furthermore, when the target parameter identification mode is in static mode, the identification of the parameters to be identified for the battery BMS system is stopped. In this mode, since the battery BMS system is not operating, parameter identification is not performed.

[0111] In this embodiment of the invention, the FFRLS algorithm employs different forgetting factors and update frequencies under different modes to optimize parameter tracking performance and improve result stability and accuracy. Furthermore, mode partitioning avoids forced identification under inapplicable conditions, thereby improving overall identification accuracy and robustness.

[0112] (V) This invention also provides a specific implementation method for identifying the parameters to be identified in the battery BMS system by using an online parameter identification algorithm, based on an updated forgetting factor, and combined with the target discretized equivalent circuit model corresponding to the battery BMS system, to obtain the parameter identification result, including:

[0113] Step S4.1: Construct a data vector based on the target discretized equivalent circuit model. The parameters in the data are state variables associated with the parameters to be identified in the battery BMS system. The state variables are determined based on the current operating condition data, as detailed in the aforementioned embodiments. This embodiment of the present invention will not elaborate further on this.

[0114] Step S4.2, perform the following operations in the current recursive process using the updated online parameter identification algorithm:

[0115] Step S4.21: Based on the error covariance matrix in the previous recursive process, the updated forgetting factor in the online parameter identification algorithm, and the data vector, determine the gain matrix in the current recursive process;

[0116] Step S4.22: Based on the parameters to be identified in the previous recursive process, the gain matrix and data vector in the current recursive process, determine the parameters to be identified in the current recursive process, so as to obtain the parameter identification result in the current recursive process.

[0117] Step S4.23: Determine the error covariance matrix in the current recursion process based on the error covariance matrix in the previous recursion process, the gain matrix in the current recursion process, the forgetting factor, and the data vector.

[0118] Specifically, the recursive process is as follows:

[0119] ;

[0120] ;

[0121] ;

[0122] in, Here is the gain matrix. Let be the error covariance matrix. For data vectors, For the parameters to be identified, The output vector of the identification result, It is the identity matrix. The forgetting factor has a value range of 0.95 to 1.

[0123] For ease of understanding, see [link to relevant documentation]. Figure 3 The flowchart shown is an online parameter identification process for FFRLS. First, the measurement data vector is constructed / updated. According to the data vector and input genetic factors Calculate the gain matrix Based on the gain matrix The initial parameters to be estimated are updated to obtain the current parameters to be identified. Based on the initial covariance matrix and genetic factors The calculated gain matrix Update the covariance matrix Repeat this process until the iteration conditions are met to obtain the identification result. It should be noted that each switch of parameter identification mode will affect the covariance matrix. Resets are performed using the values ​​given in the initial phase. (That is, the initial covariance matrix) to quickly adapt to new working conditions.

[0124] In practical applications, the FFRLS algorithm can be embedded into the battery BMS system to update the identification parameters of the battery BMS system online based on the current and voltage data collected in real time by the microcontroller. , , , , The identification results are then applied to battery state estimation (such as SOC, SOH, etc.) and battery management strategies to improve the accuracy of battery management.

[0125] This application also provides another embodiment. Traditional FFRLS, due to improper initial values ​​of the covariance matrix, is prone to slow parameter convergence and estimation divergence in the initial stage, resulting in large parameter errors and affecting subsequent battery SOX estimation. Specifically, historical operating data of the battery BMS system is identified to obtain initial identification parameters. These initial identification parameters are used as the initial values ​​for the online parameter identification algorithm to be updated. Based on the initial values, the covariance matrix after the online parameter identification algorithm has run is obtained. When the fluctuation value of the covariance matrix meets a preset condition and the online parameter identification algorithm converges, the gain matrix and the parameters to be identified in the online parameter identification algorithm are updated using the covariance matrix to obtain the online parameter identification algorithm.

[0126] In this approach, multiple resistors and capacitors identified using PSO (Proof-of-Stake) – R0, R1, R2, C1, and C2 – can be used as initial values ​​for subsequent battery parameter identification algorithms, replacing the traditional random initial value scheme. During parameter identification, the initial identified parameters serve as the initial values ​​for the FFRLS algorithm. Simultaneously, the battery voltage and real-time current collected by the microcontroller are input into the FFRLS algorithm to update the battery parameters online. When the error covariance matrix approaches a stable value (i.e., the fluctuation value of the error covariance matrix meets a preset condition), the FFRLS algorithm is considered to have converged, yielding the battery parameter identification algorithm for subsequent parameter identification. The usable data is stored for the next system power-on. For example, the adjacent errors (i.e., fluctuation values) of the covariance in the error covariance matrix within 10 sampling points are recorded. If all adjacent errors approach 0 (as a specific example of a stable value or preset condition), the algorithm is confirmed to have converged, and the battery parameters R0, R1, R2, C1, and C2 are output.

[0127] This invention, through the introduction of the FFRLS online parameter identification method and combined with a multi-mode adaptive identification mechanism, intelligently tracks the battery state throughout the entire battery lifecycle, identifies battery parameters, improves model accuracy, and provides accurate and effective battery parameters for subsequent battery SOX estimation in the BMS system. Specifically: (1) through the forgetting factor (1) Reassign the weights of historical data and new data, reduce the influence of old data on new data, and increase the weight of new data, so as to be able to track the time-varying characteristics of battery parameters; (2) By monitoring the collected current, voltage and other data, the real-time state of the battery is divided into different parameter identification modes. According to the different parameter identification modes, different parameter matching FFRLS methods are used to improve the accuracy of battery parameter identification and reduce the computational cost; (3) The recursive calculation method is adopted, which does not require storing a large amount of historical data, has a small computational load, and is suitable for embedded systems.

[0128] For ease of understanding, embodiments of the present invention also provide, as follows: Figure 4 The flowchart shown below illustrates another method for online identification of battery parameters, including:

[0129] (a) Establish a second-order RC equivalent circuit model.

[0130] (ii) Model discretization.

[0131] (III) Constructing the FFRLS algorithm framework.

[0132] (iv) Predicting working conditions in future periods using neural networks.

[0133] (v) Multi-mode adaptive identification mechanism, including: when the battery resting time exceeds a set threshold of 2 minutes, determining to enter the resting mode, not performing parameter identification and skipping FFRLS calculation; when the current data is continuous and varies within the fluctuation range of 2A, determining to enter the quasi-steady-state mode and setting a forgetting factor. Furthermore, the data is updated every 10 sampling periods; if the current data persists and the rate of change exceeds 2A, the system enters transient mode and sets a forgetting factor. 85 and it is updated once in each sampling period; when the voltage and current are 0, it is determined to enter the invalid data mode, the parameter update is paused, and the parameters of the previous moment are used.

[0134] (vi) Reset the covariant matrix during mode switching .

[0135] (vii) Perform the FFRLS algorithm for recursive calculation.

[0136] (viii) Obtain the parameters to be identified by analysis , , , , The identification results.

[0137] (ix) Parameter application: SOC / SOH estimation or battery management strategy, and return to execution (iv).

[0138] The embodiments of the present invention have at least the following features:

[0139] (1) Real-time performance and adaptive capability: The embodiment of the present invention adopts the FFRLS algorithm, which can update battery parameters in real time and adapt to parameter changes in different operating conditions, temperatures and aging states of the battery. Compared with offline methods, the embodiment of the present invention can track parameter changes online and improve the accuracy of battery state estimation; (2) High computational efficiency: The FFRLS algorithm adopts recursive calculation. Each update only requires the data at the current time and the covariance matrix at the previous time. The computational load is small and it is suitable for embedding into BMS with limited computing resources; (3) Improved accuracy: Through the forgetting factor mechanism, the influence of old data is weakened and the role of new data is emphasized, so that the parameter identification results can better reflect the current state of the battery and improve the accuracy of parameter identification; (4) Simple engineering application: The method provided by the embodiment of the present invention is easy to implement on existing BMS. Only the corresponding algorithm module needs to be added to realize the online identification of battery parameters without changing the hardware structure, which reduces the complexity of engineering application.

[0140] Based on the foregoing embodiments, this invention provides an online battery parameter identification device, see [link to previous embodiment]. Figure 5 The diagram shows a structural schematic of an online battery parameter identification device, which mainly includes the following parts:

[0141] The operating condition prediction module 502 is used to predict the future operating condition data of the battery BMS system based on the historical operating condition data of the battery BMS system.

[0142] The mode determination module 504 is used to determine the target parameter identification mode based on the future operating condition data of the battery BMS system.

[0143] The mode switching and parameter update module 506 is used to update the parameters of the online parameter identification algorithm by using the benchmark parameters corresponding to the target parameter identification mode when the switch from the current parameter identification mode to the target parameter identification mode is triggered based on future working condition data. The parameters of the online parameter identification algorithm include at least a forgetting factor.

[0144] The parameter identification module 508 is used to identify the parameters to be identified in the battery BMS system by using an online parameter identification algorithm, based on the updated forgetting factor and combined with the target discretized equivalent circuit model corresponding to the battery BMS system, to obtain the parameter identification results.

[0145] The online battery parameter identification device provided in this invention predicts future operating data based on historical operating data of the battery BMS system, determines the target parameter identification mode corresponding to the future operating data, updates the parameters of the online parameter identification algorithm using the benchmark parameters corresponding to the target parameter identification model, and identifies the parameters to be identified by combining the target discretized equivalent circuit model of the battery BMS system. This effectively improves the accuracy of parameter identification and provides accurate and effective battery parameters for subsequent battery SOX estimation in the battery BMS system.

[0146] In one implementation, the operating condition prediction module 502 is specifically used for:

[0147] Based on a pre-trained operating condition prediction model, and using historical operating condition data of the battery BMS system, the future operating condition data of the battery BMS system is predicted.

[0148] In one implementation, future operating condition data includes current data, voltage data, and battery idle time; the mode determination module 504 is specifically used for:

[0149] Match one or more of the current data, voltage data, and battery rest time with the trigger conditions of multiple pre-configured parameter identification modes;

[0150] The matched parameter identification pattern is determined as the target parameter identification pattern.

[0151] In one implementation, the parameter identification modes include a static mode, a quasi-steady-state mode, a transient mode, and an invalid data mode; the mode determination module 504 is specifically used for:

[0152] If the current data is a specified value and the battery resting time exceeds a preset time threshold, it is determined that the triggering conditions match the resting mode.

[0153] If the rate of change of the current data is within the preset fluctuation range, determine the matching with the triggering conditions of the quasi-steady-state mode;

[0154] If the rate of change of the current data exceeds the preset fluctuation range, determine the matching with the triggering conditions of the transient mode;

[0155] In the event of abnormal current data and / or abnormal voltage data, determine whether the triggering conditions match those of the standby mode.

[0156] In one implementation, the mode switching and parameter update module 506 has the function of:

[0157] When the target parameter identification mode is a quasi-steady-state mode, the forgetting factor in the online parameter identification algorithm is set to the first value, and the update frequency in the online parameter identification algorithm is set to the first update frequency.

[0158] When the target parameter identification mode is transient mode, the forgetting factor in the online parameter identification algorithm is set to a second value, and the update frequency in the online parameter identification algorithm is set to a second update frequency; wherein, the second value is lower than the first value, and the second update frequency is higher than the first update frequency;

[0159] When the target parameter identification mode is the backup mode, the forgetting factor and update frequency in the online parameter identification algorithm remain unchanged.

[0160] In one implementation, the mode switching and parameter update module 506 is further configured to:

[0161] When the target parameter identification mode is set to static mode, the identification of the parameters to be identified in the battery BMS system is stopped.

[0162] In one implementation, the parameter identification module 508 is specifically used for:

[0163] A data vector is constructed based on the target discretized equivalent circuit model corresponding to the battery BMS system. The parameters in the data vector are state variables associated with the parameters to be identified in the battery BMS system. The state variables are determined based on the current operating condition data.

[0164] The updated online parameter identification algorithm performs the following operations during the current recursive process:

[0165] Based on the error covariance matrix from the previous recursive process, the updated forgetting factor and data vector from the online parameter identification algorithm, the gain matrix in the current recursive process is determined.

[0166] Based on the parameters to be identified in the previous recursive process, the gain matrix and data vector in the current recursive process, the parameters to be identified in the current recursive process are determined, so as to obtain the parameter identification results in the current recursive process.

[0167] The error covariance matrix in the current recursion process is determined based on the error covariance matrix from the previous recursion process, the gain matrix, the forgetting factor, and the data vector.

[0168] In one implementation, the parameter identification module 508 is further configured to:

[0169] Construct an initial equivalent circuit model corresponding to the battery BMS system, and discretize the initial equivalent circuit model to obtain an initial discretized equivalent circuit model.

[0170] The initial discretized equivalent circuit model is subjected to Laplace transform to obtain the frequency domain expression of the transfer function.

[0171] The frequency domain expression transfer function is subjected to bilinear transformation, and the target discretized equivalent circuit model is obtained by combining the bilinearly transformed frequency domain expression transfer function with the initial equivalent circuit model.

[0172] The device provided in this embodiment of the invention has the same implementation principle and technical effect as the aforementioned method embodiment. For the sake of brevity, any parts not mentioned in the device embodiment can be referred to the corresponding content in the aforementioned method embodiment.

[0173] This invention provides an electronic device, specifically, the electronic device includes a processor and a memory; the memory stores a computer program, which, when run by the processor, executes the method described in any of the above embodiments.

[0174] Figure 6 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. The electronic device 100 includes: a processor 60, a memory 61, a bus 62, and a communication interface 63. The processor 60, the communication interface 63, and the memory 61 are connected through the bus 62. The processor 60 is used to execute executable modules, such as computer programs, stored in the memory 61.

[0175] The memory 61 may include high-speed random access memory (RAM) or non-volatile memory, such as at least one disk storage device. Communication between this system network element and at least one other network element is achieved through at least one communication interface 63 (which can be wired or wireless), such as the Internet, wide area network, local area network, metropolitan area network, etc.

[0176] Bus 62 can be an ISA bus, PCI bus, or EISA bus, etc. The bus can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 6 The symbol is represented by a single double-headed arrow, but this does not mean that there is only one bus or one type of bus.

[0177] The memory 61 is used to store programs. After receiving an execution instruction, the processor 60 executes the program. The method executed by the device for defining the flow process disclosed in any of the foregoing embodiments of the present invention can be applied to the processor 60 or implemented by the processor 60.

[0178] Processor 60 may be an integrated circuit chip with signal processing capabilities. In implementation, each step of the above method can be completed by the integrated logic circuitry in the hardware of processor 60 or by instructions in software form. Processor 60 can be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), etc.; it can also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field-Programmable Gate Array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this invention. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the methods disclosed in the embodiments of this invention can be directly embodied in the execution of a hardware decoding processor, or executed by a combination of hardware and software modules in the decoding processor. The software modules can reside in random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other mature storage media in the art. The storage medium is located in memory 61. Processor 60 reads the information in memory 61 and, in conjunction with its hardware, completes the steps of the above method.

[0179] The computer program product of the readable storage medium provided in the embodiments of the present invention includes a computer-readable storage medium storing program code. The instructions included in the program code can be used to execute the methods described in the foregoing method embodiments. For specific implementation, please refer to the foregoing method embodiments, which will not be repeated here.

[0180] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0181] Finally, it should be noted that the above-described embodiments are merely specific implementations of the present invention, used to illustrate the technical solutions of the present invention, and not to limit it. The scope of protection of the present invention is not limited thereto. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments within the technical scope disclosed in the present invention, or make equivalent substitutions for some of the technical features; and these modifications, changes, or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for online identification of battery parameters, characterized in that, include: Based on the historical operating data of the battery BMS system, predict the future operating data of the battery BMS system; Based on the future operating condition data of the battery BMS system, the target parameter identification mode is determined. When the future operating condition data triggers a switch from the current parameter identification mode to the target parameter identification mode, the parameters of the online parameter identification algorithm are updated using the benchmark parameters corresponding to the target parameter identification mode. The parameters of the online parameter identification algorithm include at least a forgetting factor. The future operating condition data includes current data, voltage data, and battery resting time. The parameter identification modes include resting mode, quasi-steady-state mode, transient mode, and standby mode. The resting mode is a mode where the current data is 0 and the battery resting time exceeds a preset duration threshold. The quasi-steady-state mode is a mode where the rate of change of the current data is within a preset fluctuation range. The transient mode is a mode where the rate of change of the current data exceeds the preset fluctuation range. The standby mode is a mode where the current data is abnormal and / or the voltage data is abnormal. The online parameter identification algorithm identifies the parameters of the battery BMS system based on the updated forgetting factor and the target discretized equivalent circuit model corresponding to the battery BMS system, and obtains the parameter identification results.

2. The online battery parameter identification method according to claim 1, characterized in that, Based on historical operating data of the battery BMS system, predict future operating data of the battery BMS system, including: Based on a pre-trained operating condition prediction model and historical operating condition data of the battery BMS system, the future operating condition data of the battery BMS system are predicted.

3. The online battery parameter identification method according to claim 1, characterized in that, Based on the future operating condition data of the battery BMS system, a target parameter identification mode is determined, including: Match one or more of the current data, the voltage data, and the battery resting time with the trigger conditions of a pre-configured plurality of parameter identification modes; The matched parameter identification pattern is determined as the target parameter identification pattern.

4. The online battery parameter identification method according to claim 3, characterized in that, Matching one or more of the current data, the voltage data, and the battery resting time with the trigger conditions of a pre-configured plurality of parameter identification modes, including: If the current data is a specified value and the battery resting time exceeds a preset time threshold, it is determined that the triggering conditions of the resting mode are matched. If the rate of change of the current data is within a preset fluctuation range, the triggering condition for the quasi-steady-state mode is determined to be matched. If the rate of change of the current data exceeds the preset fluctuation range, it is determined that it matches the triggering condition of the transient mode; If the current data is abnormal and / or the voltage data is abnormal, determine that the triggering conditions match the standby mode.

5. The online battery parameter identification method according to claim 4, characterized in that, The parameters of the online parameter identification algorithm are updated using the baseline parameters corresponding to the target parameter identification pattern, including: When the target parameter identification mode is the quasi-steady-state mode, the forgetting factor in the online parameter identification algorithm is set to a first value, and the update frequency in the online parameter identification algorithm is set to a first update frequency; When the target parameter identification mode is the transient mode, the forgetting factor in the online parameter identification algorithm is set to a second value, and the update frequency in the online parameter identification algorithm is set to a second update frequency; wherein, the second value is lower than the first value, and the second update frequency is higher than the first update frequency; When the target parameter identification mode is the backup mode, the forgetting factor and the update frequency in the online parameter identification algorithm remain unchanged.

6. The online battery parameter identification method according to claim 4, characterized in that, The method further includes: When the target parameter identification mode is the static mode, the identification of the parameters to be identified in the battery BMS system is stopped.

7. The online battery parameter identification method according to claim 1, characterized in that, Using the online parameter identification algorithm, based on the updated forgetting factor and combined with the target discretized equivalent circuit model corresponding to the battery BMS system, the parameters to be identified in the battery BMS system are identified to obtain parameter identification results, including: A data vector is constructed based on the target discretized equivalent circuit model corresponding to the battery BMS system. The parameters in the data vector are state variables associated with the parameters to be identified in the battery BMS system. The state variables are determined based on the current operating condition data. The updated online parameter identification algorithm performs the following operations during the current recursive process: Based on the error covariance matrix of the previous recursive process, the updated forgetting factor in the online parameter identification algorithm, and the data vector, the gain matrix of the current recursive process is determined. Based on the parameters to be identified in the previous recursive process, the gain matrix in the current recursive process, and the data vector, the parameters to be identified in the current recursive process are determined to obtain the parameter identification result in the current recursive process. The error covariance matrix in the current recursive process is determined based on the error covariance matrix from the previous recursive process, the gain matrix in the current recursive process, the forgetting factor, and the data vector.

8. The online battery parameter identification method according to claim 1 or 7, characterized in that, The method further includes: An initial equivalent circuit model corresponding to the battery BMS system is constructed, and the initial equivalent circuit model is discretized to obtain an initial discretized equivalent circuit model. The initial discretized equivalent circuit model is subjected to Laplace transform to obtain the frequency domain expression transfer function; The frequency domain expression transfer function is subjected to bilinear transformation, and the target discretized equivalent circuit model is obtained by combining the bilinearly transformed frequency domain expression transfer function with the initial equivalent circuit model.

9. A battery parameter online identification device, characterized in that, include: The operating condition prediction module is used to predict the future operating condition data of the battery BMS system based on the historical operating condition data of the battery BMS system. The mode determination module is used to determine the target parameter identification mode based on the future operating condition data of the battery BMS system. The mode switching and parameter update module is used to update the parameters of the online parameter identification algorithm using the benchmark parameters corresponding to the target parameter identification mode when switching from the current parameter identification mode to the target parameter identification mode is triggered based on the future operating condition data. The parameters of the online parameter identification algorithm include at least a forgetting factor. The future operating condition data includes current data, voltage data, and battery resting time. The parameter identification modes include resting mode, quasi-steady-state mode, transient mode, and standby mode. The resting mode is the mode where the current data is 0 and the battery resting time exceeds a preset duration threshold. The quasi-steady-state mode is the mode where the rate of change of the current data is within a preset fluctuation range. The transient mode is the mode where the rate of change of the current data exceeds the preset fluctuation range. The standby mode is the mode where the current data is abnormal and / or the voltage data is abnormal. The parameter identification module is used to identify the parameters to be identified in the battery BMS system by means of the online parameter identification algorithm, based on the updated forgetting factor and combined with the target discretized equivalent circuit model corresponding to the battery BMS system, to obtain the parameter identification result.

10. An electronic device, characterized in that, The method includes a processor and a memory, the memory storing computer-executable instructions executable by the processor, the processor executing the computer-executable instructions to implement the method of any one of claims 1 to 8.

Images