Self-adaptive active equalization method, device and system for battery management system

Abstract

The invention provides a self-adaptive active equalization method, device and system for a battery management system, and relates to the technical field of energy storage batteries, and the method comprises the steps: obtaining electrochemical impedance spectroscopy test data and charge and discharge curve segments corresponding to each battery cell in a battery pack; under the condition that the current voltage difference value of the battery cell is determined to be smaller than or equal to a preset voltage difference threshold value, inputting the electrochemical impedance spectroscopy test data and the charge and discharge curve segments into a prediction balance model to obtain a battery cell capacity attenuation rate prediction value output by the prediction balance model, and determining the capacity attenuation rate of the battery cell according to the battery cell capacity attenuation rate prediction value. A battery equalization priority ranking result is obtained; and executing corresponding energy transfer operation on the target battery cell according to the battery cell capacity fading rate predicted value and the battery equalization priority ranking result. According to the invention, accurate evaluation and efficient balance management of cell aging are realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of energy storage battery, and particularly relates to a battery management system adaptive active balancing method, device and system. BACKGROUND

[0002] The energy storage lithium battery pack is generally composed of one or several battery groups in parallel, and each battery group is composed of a plurality of batteries in series. During the charging and discharging of the lithium battery, the battery cells are prone to unbalanced short board effect, that is, when one battery is fully charged or discharged, the overall charging or discharging will stop. The traditional balancing mode is divided into active balancing and passive balancing. The active balancing adopts energy transfer mode, commonly adopts DC / DC topology structure, controls the opening and closing by monitoring the external environment (such as temperature, humidity) and single battery voltage, and fixes the current balancing to improve the charging and discharging efficiency of the battery group and save the electric energy.

[0003] The traditional active balancing acquires the overall voltage of the battery group and the voltage of each single battery cell, then sets the active balancing strategy of each single battery cell according to the acquired voltage, and controls the charging or discharging of each single battery cell according to the strategy, and then monitors the external environment during the active balancing process of each single battery cell, and adjusts the balancing process according to the result.

[0004] However, in the existing lithium battery charging and discharging process, due to the large capacity of the energy storage battery, the passive balancing current is small, and the balancing time is long. The active balancing strategy adopted depends on the voltage difference trigger, and the change of the electrochemical characteristic parameter under the aging state of the battery cell is not fully considered, and the balancing decision deviates from the actual aging state. Moreover, the traditional active balancing technology does not change the balancing current according to the aging state, and the fixed balancing current will accelerate the aging speed of the relatively aged battery cell, and further accelerate the aging difference of the whole battery group. Therefore, there is an urgent need for a battery management system adaptive active balancing method, device and system to solve the above problems. SUMMARY

[0005] In view of the problems existing in the prior art, the present application provides a battery management system adaptive active balancing method, device and system.

[0006] The present application provides a battery management system adaptive active balancing method, comprising: obtaining electrochemical impedance spectrum test data and a charge-discharge curve segment corresponding to each battery cell in the battery group; In the case where the voltage difference value of the battery cell is determined to be less than or equal to the preset voltage difference threshold, the electrochemical impedance spectrum test data and the charge-discharge curve segment are input into a prediction balancing model to obtain a battery cell capacity attenuation rate prediction value output by the prediction balancing model, and a battery balancing priority ranking result is obtained according to the battery cell capacity attenuation rate prediction value; According to the battery capacity attenuation rate prediction value and the battery equalization priority ranking result, a corresponding energy transfer operation is performed on a target battery cell in the battery pack.

[0007] According to the battery management system adaptive active equalization method provided by the application, the electrochemical impedance spectrum test data corresponding to each battery cell in the battery pack is obtained, comprising: Performing electrochemical impedance spectrum test on the battery cell, collecting solid electrolyte interface film impedance data, charge transfer resistance data and diffusion impedance data; Based on the solid electrolyte interface film impedance data, the charge transfer resistance data and the diffusion impedance data, the electrochemical impedance spectrum test data is obtained.

[0008] According to the battery management system adaptive active equalization method provided by the application, the method further comprises: Based on the least square algorithm, the voltage difference value of the battery cell is judged; If it is determined that the voltage difference value of the battery cell is greater than the preset voltage difference threshold, the equalization current adjustment operation is performed on the battery cell based on the recursive least square method, the voltage difference value, the solid electrolyte interface film impedance data and the charge transfer resistance change amount, wherein the charge transfer resistance change amount is obtained according to the charge transfer resistance data.

[0009] According to the battery management system adaptive active equalization method provided by the application, the electrochemical impedance spectrum test data is obtained based on the solid electrolyte interface film impedance data, the charge transfer resistance data and the diffusion impedance data, comprising: According to the reciprocal of the product of the solid electrolyte interface film impedance data and the solid electrolyte interface film capacitance, the solid electrolyte interface film stability index is calculated; According to the reciprocal of the charge transfer resistance data, the charge transfer efficiency value is calculated; According to the square of the reciprocal of the diffusion impedance data, the lithium ion diffusion coefficient is calculated; Based on the solid electrolyte interface film stability index, the charge transfer efficiency value and the lithium ion diffusion coefficient, the electrochemical impedance spectrum test data is constructed.

[0010] According to the battery management system adaptive active equalization method provided by the application, the prediction equalization model is trained by the following steps: According to the historical electrochemical impedance spectrum sample data and the charge-discharge curve sample, the sample data is constructed; acquire a sample value of a capacity attenuation rate of the battery cell corresponding to the historical time point of the sample data, and construct label data of the sample data based on the sample value of the capacity attenuation rate of the battery cell; train a long short-term memory neural network based on the sample data and the label data, to obtain the prediction equalization model.

[0011] According to the battery management system adaptive active equalization method provided by the application, the energy transfer operation corresponding to the target battery cell is performed according to the battery equalization priority ranking result and the predicted value of the capacity attenuation rate of the battery cell. According to the battery management system adaptive active equalization method provided by the application, the energy transfer operation corresponding to the target battery cell is performed according to the battery equalization priority ranking result and the predicted value of the capacity attenuation rate of the battery cell. Based on the bidirectional DC / DC converter and the energy distribution equalization instruction, the energy transfer operation is performed on the target battery cell.

[0012] According to the battery management system adaptive active equalization method provided by the application, the method further comprises: respectively judge the comparison results between the solid electrolyte interface film stability index, the charge transfer efficiency value and the lithium ion diffusion coefficient and the respective preset threshold values; Based on each of the comparison results, the battery cell health state data of the battery cell is updated.

[0013] The application also provides a battery management system adaptive active equalization device, comprising: A test unit is configured to acquire electrochemical impedance spectrum test data and charge-discharge curve segments corresponding to each battery cell in a battery pack. A decision unit is configured to input the electrochemical impedance spectrum test data and the charge-discharge curve segments into a prediction equalization model when it is determined that the current voltage difference value of the battery cell is less than or equal to a preset voltage difference threshold value, to obtain a predicted value of a capacity attenuation rate of the battery cell output by the prediction equalization model, and to obtain a battery equalization priority ranking result according to the predicted value of the capacity attenuation rate of the battery cell. An energy transfer unit is configured to perform a corresponding energy transfer operation on a target battery cell in the battery pack according to the predicted value of the capacity attenuation rate of the battery cell and the battery equalization priority ranking result.

[0014] The application also provides a battery management system comprising the above-mentioned battery management system adaptive active equalization device.

[0015] The application provides a battery management system adaptive active balancing method, device and system, which obtains electrochemical impedance spectrum test data and charge-discharge curve segments of each battery cell in a battery pack, inputs the battery cell voltage difference value into a prediction balancing model based on neural network training when the battery cell voltage difference value meets the standard, obtains battery cell capacity attenuation rate prediction values and sorts them, and finally performs energy transfer operation on the target battery cell according to the result, so as to realize accurate aging evaluation and efficient balancing management of the battery cell. BRIEF DESCRIPTION OF DRAWINGS

[0016] In order to more clearly illustrate the technical solutions in the application or the prior art, the drawings needed to be used in the embodiments or the prior art description will be briefly introduced below. Obviously, the drawings in the following description are some embodiments of the application, and other drawings can be obtained by those skilled in the art without creative labor on the basis of these drawings.

[0017] Figure 1 A flowchart of the battery management system adaptive active balancing method provided by the application is shown in the figure. Figure 2 A schematic diagram of the overall process provided by the application is shown in the figure. Figure 3 A structural schematic diagram of the battery management system adaptive active balancing device provided by the application is shown in the figure. Figure 4 A deployment schematic diagram of the battery management system adaptive active balancing device provided by the application on the energy storage architecture is shown in the figure. Figure 5 A structural schematic diagram of the battery management system provided by the application is shown in the figure. Figure 6 A structural schematic diagram of the electronic device provided by the application is shown in the figure. DETAILED DESCRIPTION

[0018] In order to make the purpose, technical solutions and advantages of the application more clear, the technical solutions in the application will be described clearly and completely below in combination with the drawings in the application. Obviously, the described embodiments are part of the embodiments of the application, not all the embodiments. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor are within the protection scope of the application.

[0019] The traditional battery management system (BMS) equalization technology has certain limitations. Since the single capacity of the energy storage battery is relatively large (more than several hundred AH), and the passive equalization current is generally within 5A, the equalization time will be relatively long (more than several tens of hours), and the active equalization strategy is mostly dependent on voltage difference triggering. This long-time equalization and single equalization condition active equalization mode does not fully consider the electrochemical characteristic parameters of the aging state of the battery cell at that time, such as the change of the internal resistance of the battery cell and the solid electrolyte interface (SEI) impedance, so that there is a certain deviation between the equalization decision and the actual aging state. Moreover, the traditional active equalization technology does not change the equalization current size according to the aging state, and always uses a fixed equalization current size, which will further accelerate the aging speed of the already relatively aged battery cell, thereby aggravating the aging difference of the entire battery pack.

[0020] In view of the problems existing in the prior art, the present application provides a battery management system adaptive active equalization method based on multi-dimensional electrochemical characteristics. The dynamic impedance spectrum of the battery cell is monitored in real time, and multi-parameter fusion analysis is combined to optimize the energy transfer strategy, thereby realizing efficient equalization management of the single battery cell in the battery pack. In the equalization process, the present application introduces the collection and processing of high-frequency, medium-frequency and low-frequency impedance data, forms a closed-loop control mechanism, ensures the accuracy and adaptability of the equalization decision, and significantly improves the service life and overall performance of the battery pack.

[0021] Figure 1 The flowchart of the battery management system adaptive active equalization method provided by the present application is shown in Figure 1 As shown in the figure, the present application provides a battery management system adaptive active equalization method, which comprises: Step 101: Obtain the electrochemical impedance spectrum test data and charge-discharge curve segments corresponding to each battery cell in the battery pack.

[0022] In the application, firstly, a dynamic electrochemical impedance spectroscopy (EIS) test task is performed, and a speed measuring device supporting a dynamic EIS test with a direct current bias current greater than 10 A is adopted. At the end of charging, when the state of charge (SOC) of the battery is in the interval of 80% to 90%, a pseudo-electrochemical impedance spectroscopy (PEIS) test can be performed, and the frequency range is 100 kHz to 0.1 Hz; during discharging, a galvanostatic electrochemical impedance spectroscopy (GEIS) test can be dynamically performed, the bias current is 10 A, and the change trend of diffusion impedance is monitored in real time.

[0023] After the test is completed, communication is performed with the main control part through a CAN bus (Controller Area Network), and the data transmission rate can reach 500 kbps, so as to ensure that high-precision impedance data can be reliably transmitted, and finally, electrochemical impedance spectroscopy test data corresponding to each cell in the battery pack is obtained through feature extraction. At the same time, during normal charging and discharging of the battery, the charging and discharging curve segments (such as the interval of 3.8 V to 4.1 V) of each cell are collected.

[0024] In step 102, in a case where the current voltage difference value of the cell is determined to be less than or equal to a preset voltage difference threshold value, the electrochemical impedance spectroscopy test data and the charging and discharging curve segment are input into a prediction equalization model, a cell capacity attenuation rate prediction value output by the prediction equalization model is obtained, and a battery equalization priority ranking result is obtained according to the cell capacity attenuation rate prediction value.

[0025] In the application, firstly, the current voltage difference value of the cell is judged, and when the voltage difference value is less than or equal to a preset voltage difference threshold value (for example, 30 mV), the electrochemical impedance spectroscopy test data and the charging and discharging curve segment obtained in step 101 are input into a prediction equalization model.

[0026] The prediction equalization model is obtained by training a long short term memory (LSTM) model based on a large number of training samples. The input layer of the model includes EIS data (covering high-frequency, medium-frequency and low-frequency impedance) of the last 10 cycles and a charge-discharge curve segment (3.8V to 4.1V interval), the hidden layer is composed of double-layer LSTM units (each layer has 128 neurons), the output layer can predict the capacity attenuation rate of the battery cell after 72 hours (i.e. the predicted value of the capacity attenuation rate of the battery cell), and based on the predicted value of the capacity attenuation rate of the battery cell, the battery equalization priority ranking result is output, and the prediction error is less than 2%.

[0027] Optionally, the application also provides a preset prediction trigger period, and the above prediction process is run once a day. When the state of health (SOH) of the battery cell is lower than 80%, the equalization plan is triggered 48 hours in advance, so as to prepare for the equalization of the battery cell in advance.

[0028] Specifically, in the application, the voltage difference value reflects the degree of inconsistency of the voltage between the battery cells in the battery pack. When the voltage difference between the battery cells is too large, it may affect the overall performance and service life of the battery pack, and even cause safety problems. Therefore, it is necessary to first judge whether the current voltage difference value of the battery cell is less than or equal to the preset voltage difference threshold, which serves as the basis for whether to start the subsequent prediction equalization process.

[0029] The preset voltage difference threshold is determined comprehensively according to the type, specification and actual application scenario of the battery. For example, in some scenarios with high requirements for battery performance, the threshold may be set lower, such as 30mV, to ensure that the voltage difference between the battery cells is within a small range and ensure stable operation of the battery pack; and in some scenarios with balanced requirements for cost and efficiency, the threshold may be appropriately relaxed.

[0030] In the application, the voltage monitoring module in the battery management system is used to collect the voltage values of the battery cells in the battery pack in real time, and then calculate the maximum voltage difference between the battery cells. The maximum voltage difference is compared with the preset voltage difference threshold. If the maximum voltage difference is less than or equal to the preset threshold, it means that the voltage difference between the battery cells is within an acceptable range, and the next step of the prediction equalization process can be entered; if the maximum voltage difference is greater than the preset threshold, some emergency equalization measures (such as real-time equalization fast response) may need to be taken to reduce the voltage difference, and then the prediction equalization is performed after the voltage difference meets the requirements.

[0031] Electrochemical impedance spectroscopy (EIS) is a method of analyzing the electrochemical characteristics inside a battery by measuring its impedance at different frequencies. In step 101, EIS data of each cell in the battery pack is obtained through specific test equipment, which covers high, medium and low frequency impedance information and can reflect the electrochemical parameters such as ohmic resistance, charge transfer resistance and diffusion resistance inside the cell.

[0032] The charge-discharge curve records the changes of voltage, current and other parameters of the battery over time during the charging and discharging process. The invention selects the charge-discharge curve segment in the voltage range of 3.8V to 4.1V as input data, because this voltage range usually contains some key characteristics of the battery during charging and discharging, such as the polarization characteristics of the battery, the stability of the charging and discharging platform, etc., which can provide valuable information for predicting the capacity decay rate of the cell.

[0033] Further, the collected electrochemical impedance spectroscopy test data and charge-discharge curve segments are sorted and preprocessed to ensure that the format and accuracy of the data meet the requirements of the prediction balancing model. Then, these data are input into the prediction balancing model through a specific data interface. The prediction balancing model is obtained by training a neural network model based on training samples. The neural network model has strong non-linear mapping ability and can learn the complex relationship between input data and cell capacity decay rate.

[0034] During the model training phase, a large number of training samples are needed, which include EIS data, charge-discharge curve segments and corresponding actual capacity decay rates of cells under different aging states and different use conditions. The training samples are divided into training set and validation set. The training set is used to train the neural network model, and by continuously adjusting the parameters (such as weights and biases) of the model, the error between the predicted results of the model and the actual capacity decay rate is minimized. At the same time, the validation set is used to verify the trained model and evaluate the generalization ability of the model to ensure that the model can give accurate prediction results on new data.

[0035] In the present invention, the prediction balancing model outputs the capacity decay rate prediction value of the cell after 72 hours (which can be adjusted according to actual conditions) based on the input electrochemical impedance spectroscopy test data and charge-discharge curve segments. This prediction value reflects the performance degradation trend of the cell in the future period of time. Then, according to the capacity decay rate prediction value of the cell, the cells in the battery pack are sorted to determine the priority of battery balancing. Among them, the cell with higher capacity decay rate prediction value indicates that it has a higher possibility of performance decline in the future period of time, and needs to be prioritized for balancing operation to balance the performance of each cell in the battery pack and prolong the overall service life of the battery pack. For example, the capacity decay rate prediction value is sorted from high to low, and the cells at the top of the list are prioritized for energy transfer and other balancing operations.

[0036] In step 103, according to the cell capacity attenuation rate prediction value and the battery equalization priority ranking result, a corresponding energy transfer operation is performed on the target cell in the battery pack.

[0037] In the present application, the target cell (i.e. the cell to be subjected to the energy transfer operation) is determined from the cell capacity attenuation rate prediction value and the battery equalization priority ranking result obtained in step 102. Then, the energy transfer operation is realized by using a bidirectional DC / DC converter, which supports inter-cluster or inter-box energy distribution. The master part in the battery management system sends an equalization instruction to the relevant control part through the CAN bus, and the relevant control part starts the bidirectional DC / DC converter according to the instruction to transfer the energy of the high-capacity cell to the low-capacity cell at a current of 5A. This way can shorten the equalization time to one third of the traditional method, effectively reduce energy loss, and meet the demand of large-capacity battery pack for high-efficiency equalization.

[0038] Specifically, in the present application, the target cell is determined according to the cell capacity attenuation rate prediction value and the battery equalization priority ranking result. The capacity attenuation rate prediction value reflects the degree of performance decline of the cell in the future, and the equalization priority ranking clearly shows the order of the cells that need to be subjected to the equalization operation. The cell with a higher capacity attenuation rate prediction value and a higher equalization priority ranking is determined as the target cell. For example, in a battery pack, the capacity attenuation rate prediction values of multiple cells are 5%, 3%, 7% and 2% respectively. After equalization priority ranking, the cell with a capacity attenuation rate prediction value of 7% is determined as the target cell for energy transfer operation, and after its equalization is completed, the cells in the lower order are operated in turn.

[0039] The bidirectional DC / DC converter is a power electronic device that can realize the bidirectional flow of direct current. It can change the size and direction of input and output voltage and current by controlling the conduction and shutdown of the switching tube, so as to realize the bidirectional transmission of energy. In the battery equalization system, it can transfer the energy of the high-capacity cell to the low-capacity cell, realizing the energy balance between the cells in the battery pack. Compared with the traditional unidirectional energy transfer method, the bidirectional DC / DC converter has higher flexibility and efficiency. It can flexibly control the flow direction and size of energy according to the actual demand, and can better adapt to the battery equalization demand under different working conditions. At the same time, its energy conversion efficiency is high, which can effectively reduce energy loss and improve the overall energy utilization efficiency of the battery pack.

[0040] In some large battery energy storage systems, the battery pack is usually composed of multiple battery clusters. Inter-cluster energy distribution refers to transferring energy from one battery cluster to another. For example, when the cells in a certain battery cluster are not balanced due to long-term use or other reasons, energy can be transferred from other battery clusters with sufficient energy to this battery cluster through a bidirectional DC / DC converter to achieve inter-cluster energy balancing.

[0041] For a battery system composed of multiple battery boxes, inter-box energy distribution refers to transferring energy from one battery box to another. This distribution method can further expand the range of energy balancing and improve the stability and reliability of the entire battery system. For example, in a distributed battery energy storage system, battery boxes at different locations may not be consistent due to environmental temperature, usage frequency, and other factors. Inter-box energy distribution can achieve balancing of the entire system.

[0042] In the present application, according to the cell capacity decay rate prediction value and the balancing priority ranking result, combined with the current state of the battery pack (such as the voltage, current, temperature, and other parameters of each cell), specific balancing instructions are generated. The balancing instructions will clearly indicate the target cell that needs to be transferred, the direction of energy transfer (from which cell to which cell), and the size of energy transfer, etc. The balancing instructions are sent to the slave module through the CAN bus. The CAN bus has the advantages of strong real-time performance, high reliability, and strong anti-interference ability, which can ensure that the balancing instructions are accurately and timely transmitted to the slave module.

[0043] After receiving the balancing instructions, the bidirectional DC / DC converter first analyzes the instructions and extracts key information such as the identification of the target cell, the direction and size of energy transfer, etc. Then, according to the analyzed instructions, parameters such as input and output voltage and current are adjusted to achieve directional energy transfer. During the energy transfer process, the voltage and current of the cells are monitored in real time to ensure the safety and stability of the energy transfer process. If abnormal conditions (such as excessively high or low cell voltage, excessively large current, etc.) are found, the working state of the bidirectional DC / DC converter is adjusted or the energy transfer operation is stopped, and abnormal information is fed back.

[0044] Under the action of the bidirectional DC / DC converter, the energy of the high-energy cell is gradually transferred to the low-energy cell at a certain current (such as 5A). In this process, the voltage of each cell in the battery pack gradually tends to balance, and the energy difference between the cells gradually decreases.

[0045] In the present application, the energy transfer operation can end when one of the following conditions is met: first, the voltage difference between the battery cells reaches the preset balancing target value, that is, the voltage difference between each battery cell is less than or equal to a smaller threshold, indicating that the battery pack has reached a better balanced state; second, the preset balancing time is reached, in order to avoid the energy transfer process being too long to affect the normal use of the battery pack, a maximum balancing time will be set, and the energy transfer operation will be automatically stopped when the energy transfer operation reaches the time; third, abnormal situations occur, such as high temperature of the battery cell, converter failure, etc., in order to protect the safety of the battery pack and the equipment, the energy transfer operation will be stopped immediately.

[0046] The battery management system adaptive active balancing method provided by the present application obtains the electrochemical impedance spectrum test data and the charge-discharge curve segment of each battery cell in the battery pack, inputs the prediction balancing model based on neural network training when the voltage difference value of the battery cell meets the standard, obtains the capacity attenuation rate prediction value and sorts it, and finally executes the energy transfer operation on the target battery cell according to the result, realizes the accurate evaluation and efficient balancing management of the battery cell aging.

[0047] On the basis of the above-mentioned embodiments, the electrochemical impedance spectrum test data corresponding to each battery cell in the battery pack is obtained, including: Performing electrochemical impedance spectrum test on the battery cell, collecting solid electrolyte interface film impedance data, charge transfer resistance data and diffusion impedance data; Based on the solid electrolyte interface film impedance data, the charge transfer resistance data and the diffusion impedance data, the electrochemical impedance spectrum test data is obtained.

[0048] In the present application, dynamic EIS test can be supported when the direct current bias current is greater than 10A. In the actual battery use scene, the battery cell is often in a large current working state, and the test process can simulate the impedance characteristics of the battery cell under real high current load, so that the test result is closer to the actual working condition of the battery cell, and the practicality and accuracy of the test data are improved.

[0049] Specifically, pseudo-constant potential EIS test is performed at the end of charging (SOC 80% to 90%). The end of charging is a stage in which the state of the battery changes relatively complex, and the test at this time can capture the impedance characteristics of the battery cell when it is close to full charge, which is of great significance for understanding the charging performance and health status of the battery cell.

[0050] Further, the Galvanic EIS test is dynamically performed during the discharging process. The discharging process is the stage in which the battery releases energy, and by performing the test during this process, the impedance changes of the battery under different discharging states can be monitored in real time. During the discharging process, the test is performed with a bias current of 10A, which can accurately obtain the change trend of the diffusion impedance. As the discharging proceeds, the lithium ion concentration and distribution inside the battery will change, and the diffusion impedance will also change accordingly. Through dynamic testing, these changes can be captured, providing detailed data for evaluating the discharging performance of the battery.

[0051] During the above test process, the frequency range is set to 100kHz to 0.1Hz, wherein the high frequency region (around 100kHz) mainly reflects the impedance characteristics of the solid electrolyte interface film (SEI), and through the test of this frequency band, SEI impedance data (R SEI ) can be collected. SEI is a passivation film formed by the battery during the first charging and discharging process, and its impedance characteristics have an important influence on the performance and life of the battery.

[0052] The medium frequency region (frequency range in the middle part) corresponds to the charge transfer resistance data (R ct ), and the charge transfer resistance data can be collected, which reflects the difficulty of charge transfer inside the battery and is closely related to the chemical reaction activity of the battery.

[0053] The low frequency region (around 0.1Hz) is related to the diffusion impedance data (R w ), and the diffusion impedance data can be collected. Diffusion impedance reflects the diffusion ability of lithium ions inside the battery, which has a significant impact on the charging and discharging performance of the battery.

[0054] In the present application, the collected impedance spectrum data is fitted by using an equivalent circuit model. The equivalent circuit model is a model that equivalent the electrochemical process of the battery to a combination of circuit elements. By reasonably selecting and setting the parameters of the circuit elements, the impedance characteristics of the battery under different frequencies can be simulated. For example, SEI impedance, charge transfer resistance and diffusion impedance can be equivalent to resistance elements in the circuit respectively, combined with capacitors, inductors and other elements, to construct an equivalent circuit model that can accurately describe the impedance characteristics of the battery.

[0055] For example, based on the fitted SEI impedance data, the SEI stability index SSI is calculated by a specific algorithm. SSI can reflect the stability of SEI, and the stability of SEI has an important influence on the performance and life of the battery cell. For example, if the SEI is unstable, it may cause side reactions in the battery cell during charging and discharging, increase the internal resistance of the battery cell, and reduce the capacity and cycle life of the battery cell; the charge transfer efficiency CTE is calculated according to the charge transfer resistance data, and the charge transfer efficiency CTE reflects the effective degree of charge transfer inside the battery cell, and efficient charge transfer means that the chemical reaction of the battery cell can be carried out more quickly and more fully, thereby improving the charging and discharging performance of the battery cell; the lithium ion diffusion coefficient D Li is calculated using the diffusion impedance data, and the lithium ion diffusion coefficient D Li reflects the diffusion speed of lithium ions inside the battery cell, and the larger the diffusion coefficient, the easier the diffusion of lithium ions inside the battery cell, and the better the charging and discharging rate and performance of the battery cell.

[0056] Further, the calculated SEI stability index SSI, charge transfer efficiency CTE and lithium ion diffusion coefficient D Li and other parameters are used as important components of electrochemical impedance spectrum test data to update the battery cell health status file. The battery cell health status file records various performance indicators of the battery cell at different time points, and through continuous updating of the file, the health status change trend of the battery cell can be comprehensively and dynamically understood, providing a scientific basis for maintenance, management and replacement of the battery cell. For example, when it is found that the SEI stability index SSI of a certain battery cell is continuously decreasing, the charge transfer efficiency CTE is decreasing or the lithium ion diffusion coefficient D Li is significantly reduced, it can be judged that the health status of the battery cell is deteriorating, and appropriate measures need to be taken in time, such as equalization maintenance or replacement of the battery cell.

[0057] On the basis of the above-mentioned embodiments, the method further comprises: judging the current voltage difference value of the battery cell based on the least mean square algorithm; if it is determined that the current voltage difference value of the battery cell is greater than the preset voltage difference threshold value, performing an equalization current adjustment operation on the battery cell based on the recursive least square method, the voltage difference value, the solid electrolyte interface film impedance data and the charge transfer resistance change amount, wherein the charge transfer resistance change amount is obtained according to the charge transfer resistance data.

[0058] In the present application, the Least Mean Square (LMS) algorithm is an adaptive filtering algorithm that can quickly judge the current voltage difference of the battery in the battery voltage equalization scenario. This algorithm adjusts its parameters to minimize the mean square error of the error signal. When monitoring the battery voltage in real time, it compares the actual measured voltage of each battery with the preset reference voltage (or the average voltage of the battery group), and obtains the voltage difference.

[0059] When there is a voltage difference between the batteries, the LMS algorithm can quickly respond. For example, set the preset voltage difference threshold to 30mV, once the voltage difference between the batteries is greater than 30mV, the algorithm will immediately identify this abnormal situation. This is because the LMS algorithm has the characteristics of fast convergence, and its convergence time is less than 50ms, which can complete the judgment of the voltage difference in a very short time, and provide timely basis for subsequent equalization operation. This fast response capability is crucial to maintain the consistency and stability of the battery group, and can avoid the problem of overcharging or overdischarging of the battery due to excessive voltage difference, and prolong the service life of the battery.

[0060] After determining that the current voltage difference of the battery is greater than the preset voltage difference threshold, the equalization current needs to be adjusted. Recursive Least Squares (RLS) can dynamically adjust model parameters in the process of continuously obtaining new data, so as to more accurately reflect the current state of the system. In the battery equalization system, the RLS algorithm can dynamically adjust the equalization current according to the real-time voltage difference, solid electrolyte interface film impedance data and charge transfer resistance change, to realize more accurate equalization control. Specifically, the equalization current The calculation formula is: ; Among them, represents the voltage difference between the batteries. When is greater than the preset voltage difference threshold, it means that the battery group has an imbalance phenomenon, which needs to be eliminated by adjusting the equalization current in real time. SEI is a layer of passivation film formed on the battery during the first charge and discharge process, and its impedance characteristics have an important influence on the performance of the battery. reflects the hindering effect of SEI on current, and it is one of the important impedance parameters in the calculation of equalization current. The charge transfer resistance reflects the difficulty of charge transfer within the battery, is obtained according to the charge transfer resistance data, which represents the change of the charge transfer resistance at different times or in different states, and the change of the charge transfer resistance will affect the charge and discharge performance of the battery. is an adjustment coefficient for balancing the influence degree of the charge transfer resistance change amount on the equalization current, and by reasonably setting the value of , the calculation of the equalization current can be more in line with the actual situation, and the equalization effect can be improved.

[0061] The present application combines LMS algorithm and RLS algorithm, and can realize dynamic adjustment of the equalization current. With the use and state change of the battery cell, the SEI impedance and the charge transfer resistance will also change. The RLS algorithm can continuously update the calculation parameters of the equalization current according to real-time data, so that the equalization current always remains in a suitable range. This dynamic adjustment capability can ensure that the battery cell group can realize effective equalization under different working conditions, and improve the consistency and overall performance of the battery cell group. For example, during the aging process of the battery cell, the SEI impedance may increase, and the charge transfer resistance may also change. Through dynamic adjustment of the equalization current, these changes can be compensated in time, and the equalization state of the battery cell group can be maintained.

[0062] On the basis of the above-mentioned embodiments, the obtaining of the electrochemical impedance spectrum test data based on the solid electrolyte interface film impedance data, the charge transfer resistance data and the diffusion impedance data comprises: The solid electrolyte interface film stability index is calculated according to the reciprocal of the product between the solid electrolyte interface film impedance data and the solid electrolyte interface film capacitance; The charge transfer efficiency value is calculated according to the reciprocal of the charge transfer resistance data; The lithium ion diffusion coefficient is calculated according to the square of the reciprocal of the diffusion impedance data; The electrochemical impedance spectrum test data is constructed based on the solid electrolyte interface film stability index, the charge transfer efficiency value and the lithium ion diffusion coefficient.

[0063] The solid electrolyte interface film (SEI) is a passivation film formed in the charging and discharging process of the battery cell, and its impedance characteristic (R SEI ) and capacitance characteristic have important influence on the performance of the battery cell. The present application calculates the solid electrolyte interface film stability index (SSI) by using the reciprocal of the product between the solid electrolyte interface film impedance data (R SEI ) and the solid electrolyte interface film capacitance (C SEI ). Specifically, R SEI reflects the degree of hindering of SEI to the current passing, C SEI reflects the ability of SEI to store charges, and the reciprocal of the product between the two comprehensively reflects the stability degree of SEI, for example, if R SEI is large, it means that SEI hinders the current passing greatly, and the film may be thick or the performance may be poor; if C SEIThe smaller, the weaker the storage charge capacity, which may also affect the stability of the film. The smaller the reciprocal of the product may mean that the SEI is more stable, because stable SEI should have moderate impedance and capacitance characteristics, both effectively conducting current and reasonably storing charge.

[0064] The charge transfer resistance (R ct ) reflects the difficulty of charge transfer inside the battery, which is closely related to the chemical reaction activity of the battery. The present application calculates the charge transfer efficiency value (CTE) by taking the reciprocal of the charge transfer resistance data. Specifically, the smaller the R ct , the easier the charge transfer inside the battery, the higher the electrochemical reaction activity, and the higher the charge transfer efficiency. After taking the reciprocal, the larger the CTE value, the higher the charge transfer efficiency. For example, when R ct is small, it means that the charge transfer reaction on the surface of the electrode proceeds smoothly, and the battery can charge and discharge more efficiently, so the CTE value will be larger.

[0065] The diffusion impedance (R w ) reflects the diffusion ability of lithium ions inside the battery. According to the reciprocal square of the diffusion impedance data, the present application calculates the lithium ion diffusion coefficient (D Li ). Specifically, the smaller the R w , the easier the diffusion of lithium ions inside the battery, and the stronger the diffusion ability. After taking the reciprocal square of the diffusion impedance, the larger the D Li value, the larger the lithium ion diffusion coefficient. For example, when R w is small, lithium ions can diffuse more quickly in the electrode material, which is beneficial to the charging and discharging process of the battery, so the D Li value will be larger.

[0066] In an embodiment, the SSI can be used to determine whether the SEI is stable, and if not, the charging and discharging strategy may need to be adjusted to avoid further deterioration of the SEI; the CTE value can reflect the chemical reaction activity of the battery, and if the activity decreases, the working conditions of the battery may need to be optimized; the D Li may indicate the diffusion of lithium ions, and if the diffusion is blocked, the temperature or charging and discharging rate may need to be adjusted.

[0067] The electrochemical impedance spectrum test data constructed by the solid electrolyte interface film stability index (SSI), the charge transfer efficiency value (CTE), and the lithium ion diffusion coefficient (D Li ) can provide important basis for the battery management system. Long-term monitoring and analysis of these data can evaluate the health status and performance degradation of the battery. For example, if the SSI gradually decreases, the CTE gradually decreases, or the D LiThe gradual decrease may mean that the performance of the battery cell is decreasing, and maintenance or replacement is needed in time. Meanwhile, these data can also be used to optimize the charging and discharging strategy of the battery, and improve the use efficiency and life of the battery.

[0068] In the present application, the solid electrolyte interface film stability index (SSI), the charge transfer efficiency value (CTE) and the lithium ion diffusion coefficient (D Li ) respectively reflect the electrochemical performance of the battery cell from different aspects. SSI focuses on the stability of SEI, CTE reflects the efficiency of charge transfer, and D Li embodies the diffusion ability of lithium ions. Integrating these parameters with electrochemical impedance spectrum test data can more comprehensively and comprehensively describe the internal electrochemical characteristics of the battery cell.

[0069] On the basis of the above-mentioned embodiments, the prediction equilibrium model is trained by the following steps: According to the electrochemical impedance spectrum sample data and the charge and discharge curve sample at the historical moment, sample data is constructed; Obtain the sample value of the capacity attenuation rate of the battery cell corresponding to the historical moment of the sample data, and based on the sample value of the capacity attenuation rate of the battery cell, construct the label data of the sample data; Based on the sample data and the label data, the long short-term memory neural network is trained to obtain the prediction equilibrium model.

[0070] In the present application, the construction of sample data mainly depends on the electrochemical impedance spectrum (EIS) sample data and the charge and discharge curve sample at the historical moment. The EIS data can reflect the impedance characteristics of the battery cell at different frequencies, and the impedance at different frequencies (high frequency, medium frequency and low frequency) contains different electrochemical process information inside the battery cell, for example, the high frequency impedance may be related to the charge transfer process on the electrode surface, and the medium and low frequency impedance may involve the diffusion process inside the electrode material. The charge and discharge curve sample records the change of voltage, current and other parameters of the battery cell with time during the charging and discharging process, and the charge and discharge curve segment in the range of 3.8V to 4.1V contains important electrochemical behavior information of the battery cell in this voltage range.

[0071] The input layer of the prediction equilibrium model comprises EIS data (high frequency, medium frequency, and low frequency impedance) of the last 10 cycles and a charge-discharge curve segment (3.8V-4.1V interval). When constructing sample data, the EIS data of the battery in the last 10 charge-discharge cycles is collected, the impedance values at different frequencies in each cycle are taken as a data dimension, and the charge-discharge curve segment in the 3.8V-4.1V interval of each cycle is taken as another part of data. These different cycles and different types of data are integrated together to form a complete sample data. For example, for a certain battery, the high frequency, medium frequency, and low frequency impedance values in the last 10 consecutive charge-discharge cycles and the voltage-time curve data in the 3.8V-4.1V interval of each cycle are collected, and these data are arranged and combined in a certain format to form a sample data.

[0072] Meanwhile, in the present application, the sample value of the capacity attenuation rate of the battery at the historical time is needed. The capacity attenuation rate of the battery is an important indicator for measuring the degree of performance decline of the battery, which reflects the reduction of the capacity of the battery during use. The capacity attenuation rate can be obtained by testing or long-term monitoring the charge-discharge capacity of the battery and calculating the attenuation ratio at different times relative to the initial capacity. For example, after the battery is used for a period of time, the current charge-discharge capacity is accurately measured, compared with the initial capacity, and the capacity attenuation rate is calculated.

[0073] The present application constructs label data of sample data based on the sample value of the capacity attenuation rate of the battery. The label data is a kind of annotation of the result corresponding to the sample data, that is, the capacity attenuation rate of the battery at a specific historical time. The label data provides a target value for the training of the long short-term memory neural network (LSTM), so that the network can learn the relationship between the input sample data (EIS data and charge-discharge curve segment) and the output capacity attenuation rate of the battery. For example, for a sample data, the label data corresponding to the sample is the capacity attenuation rate of the battery corresponding to the sample at a certain historical time, such as 5%.

[0074] In the present application, the long short-term memory neural network (LSTM) is selected for training. LSTM can process sequence data and has the ability to remember long-term information. In the prediction of battery performance, the historical EIS data and charge-discharge curve data of the battery are sequence data, and the performance change of the battery is a long-term process. LSTM can better capture the long-term dependence in these data, thereby improving the prediction accuracy.

[0075] In the present application, the hidden layer of the prediction equalization model is composed of double-layer LSTM units (128 neurons per layer), and the double-layer LSTM structure can further enhance the learning ability of the model for complex sequence data. The 128 neurons in each layer represent 128 processing units in each LSTM layer, which can perform complex nonlinear transformation and feature extraction on the input data. Through the combination of multiple layers and multiple neurons, the model can learn deeper features and rules in the EIS data and charge-discharge curve data of the battery cell.

[0076] The output layer of the prediction equalization model can predict the capacity attenuation rate of the battery cell after 72 hours and the equalization priority ranking. In the present application, the trained LSTM model can not only predict the capacity attenuation of the battery cell in the future 72 hours, but also rank the equalization priority of the battery cell according to the prediction result. For example, the prediction equalization model may output the capacity attenuation rate of a battery cell after 72 hours as 8%, and according to the attenuation rate and the prediction results of other battery cells, the battery cell is ranked in a higher position of equalization priority.

[0077] In the present application, the training target is to make the prediction error of the model less than 2%. By continuously adjusting the parameters (such as weights and biases) of the LSTM model, the error between the predicted value (battery cell capacity attenuation rate after 72 hours) output by the model and the actual label data (real battery cell capacity attenuation rate corresponding to the historical time) under given sample data is as small as possible. When the prediction error is less than 2%, it indicates that the model has high prediction accuracy and can provide reliable basis for the equalization management of the battery cell.

[0078] Optionally, in the present application, the prediction equalization model is run once a day, and when the SOH (state of health) of the battery cell is lower than 80%, the equalization plan is triggered 48 hours in advance. If the prediction result shows that the SOH of a battery cell is lower than 80%, it indicates that the performance of the battery cell has declined to a certain extent, and the equalization plan needs to be formulated and triggered 48 hours in advance to avoid further deterioration of the performance of the battery cell and ensure the overall performance and safety of the battery pack.

[0079] On the basis of the above-mentioned embodiments, the corresponding energy transfer operation is performed on the target battery cell in the battery pack according to the battery cell capacity attenuation rate prediction value and the battery equalization priority ranking result, comprising: According to the battery cell capacity attenuation rate prediction value and the battery equalization priority ranking result, an energy allocation equalization instruction corresponding to the target battery cell is generated; Based on the bidirectional DC / DC converter and the energy allocation equalization instruction, an energy transfer operation is performed on the target battery cell.

[0080] In the present application, the capacity attenuation rate of the battery cell is a core indicator for measuring the degree of performance decline of the battery cell, reflecting the change of the amount of charge that the battery cell can store and release over time during use. For example, a battery cell with an initial capacity of 100 Ah, after a period of use, the capacity is attenuated to 80 Ah, then the capacity attenuation rate can be calculated. The higher the capacity attenuation rate, the more serious the performance decline of the battery cell, the more it needs to be balanced to restore or maintain the overall performance of the battery pack.

[0081] The battery balancing priority ranking result is the order determined after comprehensive evaluation of the state of health, performance parameters and other factors of each battery cell in the battery pack. For example, in the ranking, battery cells with lower SEI stability index, abnormal charge transfer efficiency or unsatisfactory lithium ion diffusion coefficient may be given higher priority. This ranking result helps to prioritize the battery cells that need balancing most in the case of limited resources, improving balancing efficiency and effectiveness.

[0082] Further, in combination with the predicted value of the capacity attenuation rate of the battery cell and the balancing priority ranking result, it is clear which battery cells need to be subjected to energy transfer operation, and these battery cells are the target battery cells. For example, battery cells that are in the top of the ranking and have a high capacity attenuation rate will be selected as target battery cells. According to the specific circumstances of the target battery cells, an energy allocation balancing instruction containing parameters such as balancing current size and balancing time is generated. The adjustable balancing current range is 0.1A to 5A, which will be adjusted according to the state of the target battery cell in actual operation. For example, if a single battery has a low SEI stability index value and a high charge transfer efficiency value, the balancing current needs to be reduced (e.g., select a smaller value within the range of 0.1A to 5A, such as 0.5A) to protect SEI, while considering the high charge transfer efficiency, the balancing time may be appropriately shortened or the appropriate balancing time is maintained to achieve the overall balancing effect; for some battery cells with severe capacity attenuation and poor state of health, a larger balancing current (e.g., 3A to 5A) may be selected to speed up the energy transfer speed, but other factors will also be considered to ensure safe operation.

[0083] The bidirectional DC / DC converter is a key device for realizing inter-cluster or inter-battery box energy allocation, which can change the size and direction of direct current voltage, so that energy can be transferred bidirectionally between different battery cells, different battery clusters or different battery boxes. For example, in a battery pack, when the power of a certain battery cluster is higher and the power of another battery cluster is lower, the bidirectional DC / DC converter can transfer the energy of the high-power battery cluster to the low-power battery cluster, achieving rational allocation of energy.

[0084] The bidirectional DC / DC converter can support various energy distribution modes, meet the needs in different scenarios, accurately control the transfer direction and amount of energy according to the energy distribution balance instruction, and flexibly realize the energy distribution between clusters or between boxes.

[0085] On the basis of the above-mentioned embodiments, the method further comprises: respectively judging comparison results between the solid electrolyte interface film stability index, the charge transfer efficiency value and the lithium ion diffusion coefficient and respective corresponding preset threshold values; Based on each of the comparison results, the battery cell health state data of the battery cell is updated.

[0086] In the present application, the preset threshold values include a first preset threshold value, a second preset threshold value and a third preset threshold value, which are respectively the preset threshold values corresponding to the solid electrolyte interface film stability index, the charge transfer efficiency value and the lithium ion diffusion coefficient.

[0087] When comparing the solid electrolyte interface film (SEI) stability index, the first preset threshold value can be set to 800. The SEI stability index is a quantitative index for measuring the stability of SEI, and its value reflects the structural integrity and chemical stability of SEI in the charging and discharging process of the battery. Compare the SEI stability index obtained by real-time monitoring or calculation with the first preset threshold value. If the SEI stability index is less than 800, it indicates that the stability of SEI is poor. For example, during long-term use of the battery, due to the occurrence of side reactions, SEI may gradually thicken, crack, etc., resulting in a decrease in its stability, and the stability index monitored at this time will be lower than 800. When this situation occurs, it indicates that the SEI may further deteriorate during the subsequent charging and discharging process of the battery, thereby affecting the performance and life of the battery.

[0088] When comparing the charge transfer efficiency value, the second preset threshold value can be set to 3. The charge transfer efficiency value reflects the ability of the internal charge of the battery to transfer between the electrode and the electrolyte interface, and is an important parameter for measuring the electrochemical reaction kinetics performance of the battery. In the present application, the obtained charge transfer efficiency value is compared with the second preset threshold value. If the charge transfer efficiency value is less than 3, it indicates that the charge transfer ability is insufficient. For example, when the electrode material ages, the electrode surface is contaminated, or the electrolyte composition changes, the transfer of charge at the interface will be hindered, resulting in a decrease in charge transfer efficiency. In this case, during the charging or discharging process of the battery, the charge cannot be quickly and effectively transferred between the electrode and the electrolyte, which will affect the charging and discharging efficiency and performance of the battery.

[0089] When comparing the lithium ion diffusion coefficient, the third preset threshold can be set to 0.5. The lithium ion diffusion coefficient is a physical quantity that describes the ease of diffusion of lithium ions within the electrode material, directly affecting the charging and discharging speed and performance of the battery. In the present application, the measured lithium ion diffusion coefficient is compared with the third preset threshold. When the lithium ion diffusion coefficient is less than 0.5, it indicates that the diffusion ability of lithium ions in the electrode material is limited. For example, as the number of battery usage increases, the structure of the electrode material may change, causing the diffusion channel of lithium ions to narrow or block, resulting in a decrease in the lithium ion diffusion coefficient. At this time, during the battery charging and discharging process, lithium ions cannot diffuse in the electrode material in a timely and uniform manner, and local overcharging or overdischarging may occur.

[0090] In the present application, when the SEI stability index is less than 800, due to poor SEI stability, the charge and discharge rate may need to be reduced to reduce further degradation of SEI. When updating the battery health state data, the state of poor SEI stability is marked in the relevant data record, and the measures to be taken (reducing the charge and discharge rate) are recorded. At the same time, according to the degree of decline of SEI stability, the overall health score of the battery is adjusted accordingly, for example, a certain score is reduced, to reflect the impact of SEI degradation on the health of the battery.

[0091] If the charge transfer efficiency value is less than 3, it indicates that the charge transfer ability is insufficient, and the battery should be prioritized for charging equalization. When updating the battery health state data, the battery is marked as a priority for charging equalization, and the specific situation of low charge transfer efficiency is recorded. In addition, according to the magnitude of the decrease in charge transfer efficiency, the health state of the battery is quantitatively evaluated, such as reducing the health grade or health score of the battery, to reflect the negative impact of charge transfer problems on the performance of the battery.

[0092] When the lithium ion diffusion coefficient is less than 0.5, it indicates that the diffusion ability of lithium ions is limited, and the equalization current size needs to be adjusted to avoid the risk of overcharging or overdischarging. When updating the battery health state data, the situation of low lithium ion diffusion coefficient is recorded, and the suggestion to adjust the equalization current is noted. At the same time, according to the degree of limitation of lithium ion diffusion, the health state of the battery is updated, for example, the safety factor or health indicator of the battery is reduced, to remind the user to pay attention to the possible risks of the battery during charging and discharging.

[0093] The present application compares the solid electrolyte interface film stability index, the charge transfer efficiency value and the lithium ion diffusion coefficient with the corresponding preset threshold values respectively, and updates the battery health state data according to the comparison results, which can comprehensively and accurately grasp the health status of the battery, providing a scientific basis for the management and maintenance of the battery, thereby prolonging the service life of the battery and improving the safety and reliability of the battery system.

[0094] Figure 2 The overall flowchart provided by the present application can refer to Figure 2 As shown, first, the voltage difference value Vdiff is collected, and then it is determined whether Vdiff is greater than 30 mV. If Vdiff is greater than 30 mV, it indicates that the voltage difference between the battery cells is large, and real-time balancing operation needs to be performed immediately, entering the left real-time balancing flow. If Vdiff is not greater than 30 mV, it enters the right prediction balancing stage.

[0095] In the real-time balancing flow, the LMS algorithm is started for fast response: when Vdiff is greater than 30 mV, the least mean square (LMS) algorithm is started for fast response, and then the RLS algorithm is combined to adjust the balancing current Ieq. In the present application, on the basis of the LMS algorithm fast response, the recursive least squares (RLS) algorithm is combined to further accurately adjust the balancing current Ieq. The RLS algorithm can dynamically adjust the balancing current according to the real-time voltage difference value and other parameters to achieve better balancing effect. Further, the balancing priority is planned. In this process, the balancing priority of each battery cell is planned according to the voltage difference and other related parameters of the battery cell, ensuring that the battery management system can prioritize balancing operation on the battery cell with a larger voltage difference. Finally, the battery management system outputs specific balancing plan, including the battery cell that needs to be balanced, the size of the balancing current and the balancing time, etc., so as to perform the balancing operation.

[0096] After entering the prediction balancing flow, first, the last 10 cycle EIS data and the charge-discharge curve segment are input. These data contain the performance information of the battery cell in different states, providing input for the prediction model. Then, the prediction balancing model is run. Based on historical data and machine learning algorithm, the model can predict the future capacity decay of the battery cell. Based on the input data, the model predicts the capacity decay rate after 72 hours, so as to understand the health status of the battery cell in advance, providing basis for subsequent balancing operation. Further, according to the predicted capacity decay rate, the system generates the balancing priority ranking of each battery cell, wherein the battery cell with a higher capacity decay rate will be prioritized for balancing operation to prolong the overall service life of the battery pack. Finally, the system outputs the prediction results, including the capacity decay rate and balancing priority ranking of each battery cell, so as to facilitate subsequent balancing operation and management decision. The entire process ends after completing the real-time balancing or prediction balancing operation, and the system continues to monitor the voltage difference value of the battery cell to ensure the balancing state of the battery pack.

[0097] The battery management system adaptive active balancing device provided by the present application is described below. The battery management system adaptive active balancing device described below can be referred to in correspondence with the battery management system adaptive active balancing method described above.

[0098] Figure 3The structural schematic diagram of the battery management system adaptive active balancing device provided by the present application is shown in Figure 3 The present application provides a battery management system adaptive active balancing device, which comprises a test unit 301, a decision unit 302 and an energy transfer unit 303, wherein the test unit 301 is used to obtain the electrochemical impedance spectrum test data and the charge-discharge curve segment corresponding to each cell in the battery pack; the decision unit 302 is used to input the electrochemical impedance spectrum test data and the charge-discharge curve segment into a prediction balancing model to obtain a cell capacity attenuation rate prediction value output by the prediction balancing model, and to obtain a battery balancing priority ranking result according to the cell capacity attenuation rate prediction value, under the condition that the current voltage difference value of the cell is determined to be less than or equal to a preset voltage difference threshold value; and the energy transfer unit 303 is used to perform a corresponding energy transfer operation on a target cell in the battery pack according to the cell capacity attenuation rate prediction value and the battery balancing priority ranking result.

[0099] The battery management system adaptive active balancing device provided by the present application obtains the electrochemical impedance spectrum test data and the charge-discharge curve segment of each cell in the battery pack, inputs them into a prediction balancing model based on neural network training when the cell voltage difference value meets the standard, obtains a cell capacity attenuation rate prediction value and sorts it, and finally performs an energy transfer operation on a target cell according to the result, thereby realizing accurate aging evaluation and efficient balancing management of the cell.

[0100] On the basis of the above-mentioned embodiments, the device further comprises a feature extraction unit, which is used to perform the following steps: calculating a solid electrolyte interface film stability index according to the reciprocal of the product of the solid electrolyte interface film impedance data and the solid electrolyte interface film capacitance; calculating a charge transfer efficiency value according to the reciprocal of the charge transfer resistance data; calculating a lithium ion diffusion coefficient according to the square of the reciprocal of the diffusion impedance data; and constructing the electrochemical impedance spectrum test data based on the solid electrolyte interface film stability index, the charge transfer efficiency value and the lithium ion diffusion coefficient.

[0101] Figure 4 The deployment schematic diagram of the battery management system adaptive active balancing device provided by the present application on the energy storage architecture is shown in Figure 4 The test unit can be arranged in the battery management unit (Battery Management Unit, abbreviated as BMU) corresponding to each battery module, the feature extraction unit and the decision unit can be arranged in the battery control unit (Battery Control Unit, abbreviated as BCU), and the energy transfer unit can be arranged on each active balancing board.

[0102] Figure 5A structural schematic diagram of the battery management system provided by the present application is shown in Figure 5 The present application provides a battery management system, which comprises the battery management system adaptive active balancing device 501 described in the above embodiments.

[0103] The battery management system provided by the present application obtains the electrochemical impedance spectrum test data and the charge-discharge curve segment of each battery cell in the battery pack, inputs the data into a prediction balancing model based on neural network training when the voltage difference value of the battery cell meets the standard, obtains the battery cell capacity attenuation rate prediction value and sorts it, and finally performs energy transfer operation on the target battery cell according to the result, so as to realize accurate aging evaluation and efficient balancing management of the battery cell.

[0104] The device provided by the embodiments of the present application is used to execute the above-mentioned method embodiments, and the specific process and detailed content are referred to the above-mentioned embodiments, which will not be repeated here.

[0105] Figure 6 A structural schematic diagram of the electronic device provided by the present application is shown in Figure 6 The electronic device can include a processor (Processor) 601, a communication interface (Communications Interface) 602, a memory (Memory) 603 and a communication bus 604, wherein the processor 601, the communication interface 602 and the memory 603 complete mutual communication through the communication bus 604. The processor 601 can call the logical instructions in the memory 603 to execute the battery management system adaptive active balancing method, which comprises: obtaining the electrochemical impedance spectrum test data and the charge-discharge curve segment corresponding to each battery cell in the battery pack; in the case that the current voltage difference value of the battery cell is less than or equal to the preset voltage difference threshold value, inputting the electrochemical impedance spectrum test data and the charge-discharge curve segment into the prediction balancing model to obtain the battery cell capacity attenuation rate prediction value output by the prediction balancing model, and obtaining the battery balancing priority sorting result according to the battery cell capacity attenuation rate prediction value; according to the battery cell capacity attenuation rate prediction value and the battery balancing priority sorting result, performing corresponding energy transfer operation on the target battery cell in the battery pack.

[0106] Further, the logic instructions in the memory 603 described above can be implemented in the form of software functional units and sold or used as standalone products, which can be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the present application essentially or the parts that contribute to the prior art or parts of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium, and includes a number of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present application. The aforementioned storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a magnetic disk or an optical disk, and various media that can store program codes.

[0107] In another aspect, the present application also provides a computer program product, which comprises a computer program stored on a non-transitory computer readable storage medium, the computer program comprising program instructions which, when executed by a computer, enable the computer to perform the battery management system adaptive active balancing method provided by the above-mentioned methods, the method comprising: obtaining electrochemical impedance spectrum test data and a charge-discharge curve segment corresponding to each cell in a battery pack; in a case where it is determined that a current voltage difference value of the cell is less than or equal to a preset voltage difference threshold, inputting the electrochemical impedance spectrum test data and the charge-discharge curve segment into a prediction balancing model to obtain a cell capacity attenuation rate prediction value output by the prediction balancing model, and obtaining a battery balancing priority ranking result according to the cell capacity attenuation rate prediction value; and performing a corresponding energy transfer operation on a target cell in the battery pack according to the cell capacity attenuation rate prediction value and the battery balancing priority ranking result.

[0108] In yet another aspect, the present application also provides a non-transitory computer readable storage medium having a computer program stored thereon, which, when executed by a processor, implements a battery management system adaptive active balancing method provided by the above-mentioned embodiments, the method comprising: obtaining electrochemical impedance spectrum test data and a charge-discharge curve segment corresponding to each cell in a battery pack; in a case where it is determined that a current voltage difference value of the cell is less than or equal to a preset voltage difference threshold, inputting the electrochemical impedance spectrum test data and the charge-discharge curve segment into a prediction balancing model to obtain a cell capacity attenuation rate prediction value output by the prediction balancing model, and obtaining a battery balancing priority ranking result according to the cell capacity attenuation rate prediction value; and performing a corresponding energy transfer operation on a target cell in the battery pack according to the cell capacity attenuation rate prediction value and the battery balancing priority ranking result.

[0109] The apparatus embodiments described above are merely illustrative, wherein the units described as separated components can or can not be physically separated, and the components displayed as units can or can not be physical units, i.e., can be located in one place, or can be distributed to multiple network units. Part or all of the modules can be selected to achieve the purpose of the embodiment scheme according to actual needs. Those skilled in the art can understand and implement without creative labor.

[0110] Through the description of the above embodiments, those skilled in the art can clearly understand that the embodiments can be realized by means of software and the necessary general hardware platform, and of course can also be realized by hardware. Based on such understanding, the above technical solutions can be embodied in the form of a software product, which can be stored in a computer readable storage medium, such as a ROM / RAM, a magnetic disk, an optical disk, etc., and includes a plurality of instructions to make a computer device (which can be a personal computer, a server, or a network device, etc.) execute the methods described in each embodiment or some parts of the embodiments.

[0111] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, and not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: it can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement to part of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A battery management system adaptive active balancing method, characterized in that, The method comprises the following steps: obtaining electrochemical impedance spectrum test data and charge-discharge curve segments corresponding to each battery cell in the battery pack; in a case where it is determined that the current voltage difference value of the battery cell is less than or equal to a preset voltage difference threshold, inputting the electrochemical impedance spectrum test data and the charge-discharge curve segments into a prediction equalization model to obtain a battery cell capacity attenuation rate prediction value output by the prediction equalization model, and obtaining a battery equalization priority ranking result according to the battery cell capacity attenuation rate prediction value; performing a corresponding energy transfer operation on a target battery cell in the battery pack according to the battery cell capacity attenuation rate prediction value and the battery equalization priority ranking result.

2. The battery management system adaptive active balancing method of claim 1, wherein, The method comprises the following steps: performing electrochemical impedance spectrum testing on the battery cell to collect solid electrolyte interface film impedance data, charge transfer resistance data and diffusion impedance data; obtaining the electrochemical impedance spectrum test data based on the solid electrolyte interface film impedance data, the charge transfer resistance data and the diffusion impedance data.

3. The battery management system adaptive active balancing method of claim 2, wherein, The method further comprises the following steps: judging the current voltage difference value of the battery cell based on a least mean square algorithm; in a case where it is determined that the current voltage difference value of the battery cell is greater than the preset voltage difference threshold, performing an equalization current adjustment operation on the battery cell based on a recursive least square method, the voltage difference value, the solid electrolyte interface film impedance data and a charge transfer resistance change amount, wherein the charge transfer resistance change amount is obtained according to the charge transfer resistance data.

4. The battery management system adaptive active balancing method of claim 2, wherein, The method of obtaining the electrochemical impedance spectrum test data based on the solid electrolyte interface film impedance data, the charge transfer resistance data and the diffusion impedance data comprises the following steps: calculating a solid electrolyte interface film stability index according to the reciprocal of the product of the solid electrolyte interface film impedance data and a solid electrolyte interface film capacitance; calculating a charge transfer efficiency value according to the reciprocal of the charge transfer resistance data; calculating a lithium ion diffusion coefficient according to the square of the reciprocal of the diffusion impedance data; constructing the electrochemical impedance spectrum test data based on the solid electrolyte interface film stability index, the charge transfer efficiency value and the lithium ion diffusion coefficient.

5. The battery management system adaptive active balancing method of claim 2, wherein, The prediction equalization model is trained by the following steps: constructing sample data according to historical electrochemical impedance spectrum sample data and charge-discharge curve samples; obtaining battery cell capacity attenuation rate sample values corresponding to the sample data at historical time points, and constructing label data of the sample data based on the battery cell capacity attenuation rate sample values; training a long short-term memory neural network based on the sample data and the label data to obtain the prediction equalization model.

6. The battery management system adaptive active balancing method of claim 2, wherein, The method of performing a corresponding energy transfer operation on a target battery cell in the battery pack according to the battery cell capacity attenuation rate prediction value and the battery equalization priority ranking result comprises the following steps: generating an energy allocation equalization instruction corresponding to the target battery cell according to the battery cell capacity attenuation rate prediction value and the battery equalization priority ranking result; Performing energy transfer operation on the target cell based on the bi-directional DC / DC converter and the energy distribution balancing instruction.

7. The battery management system adaptive active balancing method of claim 4, wherein, The method further comprises: Respectively judging comparison results between the solid electrolyte interface film stability index, the charge transfer efficiency value and the lithium ion diffusion coefficient and respective preset threshold values; Updating the cell health state data of the cell based on the respective comparison results.

8. A battery management system adaptive active balancing device, characterized in that, Comprise: A test unit configured to obtain electrochemical impedance spectrum test data and charge-discharge curve segments corresponding to each cell in the battery pack; A decision unit configured to, in a case where it is determined that the current voltage difference value of the cell is less than or equal to a preset voltage difference threshold value, input the electrochemical impedance spectrum test data and the charge-discharge curve segments into a prediction balancing model to obtain a cell capacity attenuation rate prediction value output by the prediction balancing model, and obtain a battery balancing priority ranking result according to the cell capacity attenuation rate prediction value; An energy transfer unit configured to perform a corresponding energy transfer operation on a target cell in the battery pack according to the cell capacity attenuation rate prediction value and the battery balancing priority ranking result.

9. A battery management system, characterized by, The battery management system adaptive active balancing device of claim 8.

Images