Lithium battery pack chip equalization control method and control system thereof

Abstract

The invention provides a lithium battery pack chip equalization control method and a control system thereof. The method comprises the following steps: acquiring battery monomer data; matching the health score of the single battery with the health level to obtain a final balance control quantity of the single battery; inputting the equalization intensity of each battery monomer and the health score of the battery monomer into an SOM network to obtain an adaptive equalization strategy of the battery monomer; the battery health prediction model is used for capturing a health change trend of a battery monomer on a long-time scale; and judging whether the battery is abnormal or not based on the error between the predicted health state of the battery and the actual health state of the battery, and obtaining a final equalization strategy of the battery. According to the invention, the problems of static equalization strategy, low-efficiency equalization control, lack of intelligent management and the like in the prior art are effectively solved, the equalization efficiency of the battery pack is improved, the service life of the battery is prolonged, and the adaptability and robustness of the system in a complex and dynamic environment are enhanced.

Description

Technical Field

[0001] The present invention belongs to the field of lithium battery pack chip balancing control, and in particular relates to a lithium battery pack chip balancing control method and a control system thereof. Background Art

[0002] With the rapid development of electric vehicles, wearable devices, large-scale energy storage systems and other fields, lithium batteries, as a high-energy-density, long-life power source option, have been widely used in many fields. In order to ensure the stability and safety of lithium batteries, the battery management system (BMS) has become a core component to ensure the performance of lithium batteries. In the battery management system, battery balancing control is a key technology, which is used to ensure the balance of power of each single cell in the battery pack and prevent some single cells from being damaged due to overcharging or over-discharging, thereby extending the service life of the battery pack and improving its overall efficiency.

[0003] At present, there are two main types of battery balancing control technologies: passive balancing and active balancing. Passive balancing technology dissipates excess electrical energy in the battery in the form of heat through a simple resistance heat release method, thereby achieving power balancing between batteries. Although this method is simple to implement and low-cost, its efficiency is low because the excess energy in the battery is directly converted into heat and wasted, resulting in serious energy loss. Active balancing technology transfers the excess energy of the single cell with higher power in the battery pack to the single cell with lower power through energy conversion, transfer or storage, which can effectively reduce energy waste and improve balancing efficiency. However, the system complexity and cost of active balancing technology are high, and it requires more sophisticated control algorithms and hardware support.

[0004] Although the existing battery balancing control methods can solve the problem of battery cell power imbalance to a certain extent, there are still several technical bottlenecks that need to be solved urgently. First, most existing technologies still rely on static balancing strategies and lack the ability to dynamically adapt to the real-time health status and usage environment of the battery pack. This means that during the operation of the battery pack, when the health status of the battery cell changes or the external environment fluctuates, the system often cannot adjust the balancing strategy in time, resulting in a decrease in balancing efficiency and may even cause safety hazards. Secondly, the existing battery balancing system generally lacks intelligent adaptive control and cannot be optimized and adjusted according to multiple factors such as the usage status, aging status, temperature changes, etc. of the battery cell. Although some advanced control algorithms (such as adaptive algorithms, machine learning, etc.) have been proposed, most of these technologies face great challenges in hardware implementation, real-time performance, and computing resource requirements. Furthermore, in some battery pack usage scenarios, such as long-term dormancy or no data communication, traditional battery balancing technology cannot effectively maintain the balancing state, resulting in battery performance degradation and affecting battery life.

[0005] In addition, the battery balancing control strategy lacks deep integration with the overall health management of the battery pack. Existing technologies are usually limited to the balancing of battery power, ignoring the combined impact of other key factors such as battery health status and charging and discharging efficiency, and cannot achieve comprehensive battery management. More importantly, the traditional balancing system fails to effectively combine advanced sensor technology, big data analysis and artificial intelligence algorithms, which makes it impossible for the system to perform efficient and accurate monitoring and management when facing large-scale, distributed battery packs. Especially in complex and dynamic working environments, the system's robustness and adaptability are poor. Summary of the invention

[0006] The purpose of the present invention is to propose a lithium battery pack chip balancing control method and control system, which can achieve comprehensive management of battery health by accurately monitoring the status data of each battery cell, and improve the adaptability and robustness of the system in complex environments by adjusting the balancing strategy in real time. Through dynamic stratification and adaptive balancing of battery cells, the balancing strategy can be intelligently adjusted according to factors such as the health status, temperature changes, and load fluctuations of the battery cells, maximizing the balancing efficiency while avoiding battery performance loss.

[0007] In order to achieve the above object, a first aspect of the present invention provides a lithium battery pack chip balancing control method, the method comprising the following steps: Obtain battery cell data, build a machine learning model based on the battery cell data to evaluate the battery health status, and obtain a health score of the corresponding battery cell; the machine learning model is trained using a feature vector after the battery cell data is fused; The health score of the battery cell is matched with the health level to determine the health level of each battery cell. Based on the health score of the battery cell, the balancing strength of each battery cell is determined by optimizing the decision model. The balancing strength of the battery cell is subjected to the first regularization operation to obtain the final balancing control amount of the battery cell. ; The balancing strength of each battery cell and the health score of the battery cell are input into the SOM network to obtain the adaptive balancing strategy of the battery cell; the adaptive balancing strategy of the battery cell is subjected to a second regularization operation based on the final balancing control amount of the battery cell to obtain the final balancing strategy The SOM network maps the battery cells into similar groups according to the balancing strength of each battery cell and the health score of the battery cell, and is obtained through unsupervised learning training; The final balancing strategy, the health score of the battery cell and the balancing strength of each battery cell are input as data sets into a battery health prediction model based on a long short-term memory network to obtain a predicted health state of the battery cell; the battery health prediction model is used to capture the health change trend of the battery cell over a long time scale; Based on the error between the predicted health status of the battery cell and the actual health status of the battery cell, it is judged whether it is abnormal. If the error exceeds the threshold, an alarm is triggered and the self-recovery mechanism is entered, and adaptive adjustments are made according to the actual situation to reduce the risk of battery damage; the self-recovery mechanism adjusts the balancing strategy based on the current health status and balancing strength of the battery cell to obtain the final balancing strategy for the battery cell. .

[0008] Further, the battery cell data includes voltage, current, temperature and internal resistance; When acquiring battery cell data, preprocessing is also performed: The weighted moving average method is used to smooth the data of each battery cell to obtain the smoothed data; Performing standardization on the smoothed data to obtain standardized data; The standardized data are fused and analyzed and combined into a feature vector.

[0009] Furthermore, the machine learning model is trained based on the feature vector to obtain a health score of the battery cell; the health score of the battery cell ranges from [0,1], where 0 indicates that the battery health state is extremely poor and 1 indicates that the battery is in the best state.

[0010] Further, the health level includes health level 1, health level 2, and health level 3; Among them, the health level 1 indicates that the health of the battery cell is the worst and balancing is required as the top priority; the health level 2 indicates that the health of the battery cell is medium and regular balancing is performed; the health level 3 indicates that the health of the battery cell is the best and the balancing requirement is the least; The optimization decision model determines the balancing strength of each battery cell and is expressed as: ; in, Indicates battery cells At the moment Balanced strength; and is the weight parameter, , flexibility in controlling the equalization intensity; For battery cells At the moment health score; The first regularization term , which is used to control excessive intervention in the equilibrium process, is expressed as: ; in, It is the first regularization term in the balancing process, which is used to prevent over-balancing and ensure that the health status of the battery is not excessively disturbed; is the regularization coefficient, which controls the degree of intervention and is usually set between between.

[0011] Furthermore, the training goal of the SOM network is to minimize the error between neurons, so that each battery cell is mapped to the neuron closest to its health state and balance strength. During the training process, the weight of each neuron will be updated according to the input information to form the corresponding battery cell characteristics; the training is expressed as: ; in, is the updated weight; To represent neurons At the moment The weight reflects the balance demand and health status of the battery cells; is the learning rate, which controls the magnitude of each update; is the neighborhood function, which indicates the influence range of each neuron; The input data includes the battery health score and equalization strength.

[0012] Furthermore, the trained SOM network generates an adaptive balancing strategy for each battery based on the health status and balancing requirements of the battery cells. , expressed as: ; in, To represent a battery cell At the moment Adaptive balancing strategy; is the neuron weight in the SOM network, reflecting the characteristics of the battery cell; Provide health scores for battery cells; It is the balanced strength of battery cells.

[0013] Furthermore, the second regularization operation is used to control the balancing operation of each battery to be not too severe, and is defined as: ; in, Indicates battery cells At the moment The second regularization term of is the regularization coefficient, which is used to adjust the range of the equilibrium strategy; Represents the ideal balancing strategy, which is usually set based on the optimal battery model.

[0014] Furthermore, the error between the predicted health state of the battery cell and the actual health state of the battery cell is measured based on an error metric function; the error metric function is expressed as: ; in, is the error measure, is the actual health status of the battery cell, The predicted health status of the battery cells.

[0015] Furthermore, the self-recovery mechanism designs a mild abnormality strategy and a severe abnormality strategy according to the abnormality degree of the battery cell; The mild abnormality strategy is to adjust the battery balancing strategy To alleviate the problem and return the battery to normal working state; the severe abnormality strategy will suspend or reduce the balancing operation of the battery cell and optimize the performance of the battery pack by redistributing the load.

[0016] In a second aspect of the present invention, a lithium battery pack chip balancing control system is provided, the system comprising: A battery cell data acquisition unit, used to acquire battery cell data, build a machine learning model based on the battery cell data to evaluate the battery health status, and obtain a health score of the corresponding battery cell; the machine learning model is trained using a feature vector after the battery cell data is fused; The battery cell health analysis unit is used to match the health score of the battery cell with the health level, determine the health level of each battery cell, determine the balancing strength of each battery cell by optimizing the decision model based on the health score of the battery cell, and perform the first regularization operation on the balancing strength of the battery cell to obtain the final balancing control amount of the battery cell. ; A strategy training unit is used to input the balancing strength of each battery cell and the health score of the battery cell into the SOM network to obtain an adaptive balancing strategy for the battery cell; and perform a second regularization operation on the adaptive balancing strategy of the battery cell based on the final balancing control amount of the battery cell to obtain a final balancing strategy. The SOM network maps the battery cells into similar groups according to the balancing strength of each battery cell and the health score of the battery cell, and is obtained through unsupervised learning training; A fault prediction unit, used to input the final balancing strategy, the health score of the battery cell and the balancing strength of each battery cell as a data set into a battery health prediction model based on a long short-term memory network to obtain a predicted health state of the battery cell; the battery health prediction model is used to capture the health change trend of the battery cell on a long time scale; The control optimization unit is used to determine whether there is an abnormality based on the error between the predicted health state of the battery cell and the actual health state of the battery cell. If the error exceeds the threshold, an alarm is triggered and the self-recovery mechanism is entered, and adaptive adjustments are made according to the actual situation to reduce the risk of battery damage; the self-recovery mechanism adjusts the balancing strategy based on the current health state and balancing strength of the battery cell to obtain the final balancing strategy for the battery cell. .

[0017] The beneficial technical effects of the present invention are at least as follows: Adaptive hierarchical battery balancing control strategy: This method first monitors the health status of each battery cell in real time (such as voltage, current, temperature, capacity, etc.), and divides the battery cells into different levels according to these parameters. Different balancing strategies are adopted for different health states of battery cells. For example, healthy batteries only require slight balancing adjustments, while batteries that are severely aged or have abnormal temperatures require more sophisticated active balancing strategies. This hierarchical control strategy not only reduces the intervention of batteries with relatively balanced charge in the battery pack, but also optimizes the management of aging batteries and improves the overall efficiency of balancing.

[0018] Self-organizing neural network (SOM) self-optimizing battery balancing algorithm: This invention combines the self-organizing neural network (SOM) algorithm to perform dynamic learning and self-optimization through real-time monitoring of the health data and working status of battery cells. SOM analyzes the similarities of battery cells through clustering, groups the batteries according to similarities, and adjusts its balancing strategy according to the characteristics of each group. This innovation effectively overcomes the static nature of traditional balancing systems and provides an adaptive, dynamically optimized balancing control method that can respond to changes in battery packs in real time and improve balancing efficiency and accuracy.

[0019] Prediction management and self-recovery mechanism: The present invention also introduces a battery health prediction algorithm based on machine learning, which combines the battery's historical data to predict the battery's future state. When the system predicts that the performance of certain battery cells will decline, it will start the corresponding balancing adjustment in advance to avoid excessive attenuation of battery performance. In addition, the self-recovery mechanism designed by the present invention can adjust the balancing strategy in time when an abnormality occurs during the battery balancing process, restore the system to normal working state, and ensure the long-term stable operation of the battery pack.

[0020] Through the above innovations, the present invention effectively solves the problems existing in the prior art such as static balancing strategies, inefficient balancing control and lack of intelligent management. It not only improves the balancing efficiency of the battery pack and extends the battery life, but also enhances the adaptability and robustness of the system in complex and dynamic environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] The present invention is further described using the accompanying drawings, but the embodiments in the accompanying drawings do not constitute any limitation to the present invention. A person skilled in the art can obtain other drawings based on the following drawings without creative work.

[0022] Figure 1 A flow chart of a lithium battery pack chip balancing control method disclosed in an embodiment of the present invention Figure 2 This is a framework diagram of a lithium battery pack chip balancing control system disclosed in an embodiment of the present invention. DETAILED DESCRIPTION

[0023] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and cannot be understood as limiting the present invention.

[0024] Embodiment 1 like Figure 1 As shown, an embodiment of the present invention provides a lithium battery pack chip balancing control method, the method comprising: S101. Obtain battery cell data, build a machine learning model based on the battery cell data to evaluate the battery health status, and obtain a health score of the corresponding battery cell; the machine learning model is trained using a feature vector after the battery cell data is fused.

[0025] Specifically, in the prior art, battery health monitoring is usually discontinuous or lacks sufficient accuracy. The battery health status is not fully evaluated based on the multi-dimensional data of the battery (such as voltage, current, temperature, etc.), resulting in inaccurate balancing control. Therefore, this solution proposes a new method to achieve accurate monitoring of battery health status through real-time, multi-dimensional data analysis and fusion.

[0026] The battery management system (BMS) collects data from battery cells in real time through a variety of sensors (such as voltage, current, temperature, internal resistance sensors, etc.). The multidimensional data is: Voltage , Current ,temperature 、Internal resistance ; These data are the basis for evaluating the battery health status. The collected data are expressed as: ; Furthermore, since the collected data may contain noise, it needs to be smoothed through signal processing. The weighted moving average (WMA) method is used to smooth each feature (such as voltage, current, etc.) to make the data more reliable. The specific formula is as follows: ; in, is the smoothed voltage data; For battery cells exist Voltage data at the moment; As the weighting coefficient, an exponential decay function is usually used: ,in is the weighting coefficient, , making the recent data weight higher. This method can eliminate the fluctuations caused by instantaneous noise and improve data quality.

[0027] Furthermore, in order to avoid the influence of features of different dimensions on model training, the data needs to be standardized. For battery cells At the moment The voltage data is standardized using the following formula: ; in, and Battery cells Minimum and maximum voltage values; is the standardized voltage data. All features are standardized in the same way to ensure that their range is between [0,1], which is helpful for the subsequent training of machine learning models.

[0028] Furthermore, since the battery health status is determined by multiple factors, it is necessary to perform fusion analysis on the multi-dimensional data of the battery. First, the standardized data of each battery is combined into a feature vector: ; The feature vector contains the standardized voltage, current, temperature and internal resistance data of each battery cell. Through the joint analysis of these multi-dimensional data, the health status of the battery can be fully reflected.

[0029] Furthermore, a health assessment model is constructed using machine learning models (such as random forest regression, support vector regression, etc.) to assess the battery health status. Based on the standardized multidimensional data, the training model predicts the battery health score. The formula is as follows: ; in, For battery cells At the moment health score; A trained machine learning model (such as random forest regression); is a parameter of the model. The health score is generated based on the real-time data of the battery and can reflect the current health status of the battery.

[0030] Furthermore, the final output is the health score of each battery cell. , which reflects the current state of the battery, usually in the range of [0,1], where 0 indicates that the battery is in a very poor health state and 1 indicates that the battery is in the best state. The health score can be used for subsequent balancing control and early warning systems.

[0031] ; in, A collection of health scores for all battery cells.

[0032] This step provides a battery health assessment method based on multi-dimensional data, which uses a machine learning model to accurately assess the battery's health status by real-time monitoring of the battery's voltage, current, temperature, internal resistance and other characteristic data, and combining data denoising, smoothing, standardization and fusion analysis. This process effectively solves the problems of discontinuity and lack of accuracy in battery health monitoring in the prior art, and can provide an accurate battery health score, providing a decision-making basis for subsequent balancing control and early warning systems.

[0033] S102, matching the health score of the battery cell with the health level to determine the health level of each battery cell, determining the balancing strength of each battery cell by optimizing the decision model based on the health score of the battery cell, and performing a first regularization operation on the balancing strength of the battery cell to obtain the final balancing control amount of the battery cell .

[0034] In the prior art, battery balancing control usually relies on a single balancing strategy, which is difficult to adjust accurately according to the actual health status of the battery. The health status of different battery cells varies greatly, which may lead to low balancing efficiency and even shorten the battery life. Therefore, this step proposes a hierarchical management method based on battery health status scoring, which optimizes the balancing process and improves the overall balancing efficiency through hierarchical processing and differentiated control strategies.

[0035] Furthermore, the battery health status is stratified: According to each battery cell obtained in step S101 Health score , all battery cells are divided into different health levels. The specific stratification criteria are set according to the range of battery health scores: Health Level 1 ( ): Indicates that the health of the battery cells is poor and balancing is required as a priority.

[0036] Health Level 2 ( ): Indicates that the health of the battery cells is moderate and regular balancing can be performed.

[0037] Health Level 3 ( ): Indicates that the health of the battery cells is good and the balancing requirement is low.

[0038] This hierarchical approach allows for a personalized balancing strategy tailored to the health of each battery cell.

[0039] Furthermore, different balancing control strategies are designed for batteries at different health levels. Specifically: Health Level 1 ( ): For batteries in poor health, strong equalization is required, using high-speed discharge and charge regulation to restore the battery's energy balance as quickly as possible. The current can be increased gradually to ensure rapid stabilization of the battery.

[0040] Health Level 2 ( ): For batteries in medium health, a medium-speed balancing strategy is used to gradually bring the battery closer to the ideal state through fine current regulation.

[0041] Health Level 3 ( ): For batteries in good health, a low-speed balancing strategy is used to reduce interference to the batteries during the balancing process, and balancing is performed only when there is a significant difference in power between the batteries.

[0042] Furthermore, an optimization decision model is introduced to optimize the health score of each battery cell. , determine the equilibrium strength of each monomer The output of the decision model is the balancing strategy and adjustment amount for each battery cell. In order to dynamically adjust the balancing strategy, a weighted optimization model based on health score is designed: ; in, Indicates battery cells At the moment Balanced strength; and is the weight parameter, , flexibility in controlling the equalization intensity; For battery cells At the moment The formula shows that the balanced strength With battery health score By adjusting the weight and , the balancing strategy can be flexibly adjusted according to actual needs.

[0043] Furthermore, based on the above-mentioned balanced strength , balancing is performed by adjusting the current and the charge and discharge rate. In order to further improve the balancing efficiency, an additional first regularization term is introduced , used to control excessive intervention in the equilibrium process: ; in, It is the first regularization term in the balancing process, which is used to prevent over-balancing and ensure that the health status of the battery is not excessively disturbed; is the regularization coefficient, which controls the degree of intervention and is usually set between between.

[0044] The final equilibrium decision and execution is made by weighting the equilibrium strength and the regularization term To determine, so as to ensure the balancing effect while avoiding unnecessary damage to the battery.

[0045] ;

[0046] in, For battery cells The final balance control quantity.

[0047] This step designs differentiated balancing control strategies based on the hierarchical management method of battery health status to adapt to the health status of different battery cells. Through the weighted optimization model and regularization method, a dynamic and intelligent balancing process can be achieved, which significantly improves the balancing efficiency and effectively avoids excessive intervention in the battery. This method can significantly optimize the energy management of the battery pack, increase the service life of the battery, and solve the problem of insufficient refinement and low efficiency of balancing control in the prior art.

[0048] S103, inputting the balancing strength of each battery cell and the health score of the battery cell into the SOM network to obtain an adaptive balancing strategy for the battery cell; performing a second regularization operation on the adaptive balancing strategy of the battery cell based on the final balancing control amount of the battery cell to obtain a final balancing strategy The SOM network maps battery cells into similar groups according to the balancing strength of each battery cell and the health score of the battery cell, and is trained through unsupervised learning.

[0049] Specifically, the input of this step is the output of step S102: that is, the battery stratification data generated by the battery health status and equalization strength in step S102 and , these data are at every moment Reflects the health of the battery cells and their balancing needs.

[0050] In this step, the SOM network maps these input data into a high-dimensional space and clusters the characteristics of the battery cells through unsupervised learning methods. This step is critical because the present invention no longer simply adjusts the balance according to fixed rules, but through the adaptive ability of SOM, the network can dynamically adjust the strategy according to the health changes of the battery. The specific input data is: ; This data provides the SOM with sufficient input information to enable it to respond quickly to the health status of each battery cell.

[0051] Furthermore, the self-organizing neural network (SOM network) is trained through unsupervised learning, and the core goal is to map the battery cells into similar groups according to the balancing needs and health status of each battery cell. This process is based on the competitive learning mechanism of SOM, and each battery cell finds its most suitable category in the SOM network.

[0052] The training goal of SOM is to minimize the error between neurons. Specifically, the goal is to map each battery cell to the neuron closest to its health status and balance strength. In this process, the weight of each neuron will be updated according to the input information to form the corresponding battery cell characteristics.

[0053] The training formula is: ; in, Representing neurons At the moment The weight reflects the balance demand and health status of the battery cells; is the learning rate, which controls the magnitude of each update; is the neighborhood function, which indicates the influence range of each neuron; The input data includes the battery health score and balancing strength. This training process ensures that the neural network can cluster according to the real-time health status of the battery, and as the training progresses, the network's adaptability to different battery cells gradually increases, ensuring the adaptability of the balancing strategy.

[0054] Furthermore, the trained SOM network generates the optimal balancing strategy for each battery based on the health status and balancing requirements of the battery cells. In practical applications, the balancing strategy of a battery is usually expressed as its charging or discharging rate. , which is the operating intensity required by each battery cell at a specific moment.

[0055] The innovation of this step is that the adaptive strategy generation based on SOM can be dynamically adjusted according to the actual health status and balancing requirements of the battery cells. In order to enable each battery cell to obtain the most suitable balancing strength , the present invention defines a mapping function: ; in, To represent a battery cell At the moment Adaptive balancing strategy; is the neuron weight in the SOM network, reflecting the characteristics of the battery cell; Provide health scores for battery cells; is the balancing strength of the battery cell. This function combines the weight information of the SOM network training , can accurately generate a balancing strategy suitable for the current state of the battery cells.

[0056] Furthermore, in order to avoid excessive interference with the normal operation of the battery cell, this step introduces a second regularization term , which aims to control the balancing operation of each battery from being too intense. The regularization term can not only control the range of the balancing strategy, but also reduce the repeated adjustments of the balancing operation to avoid excessive interference with the normal discharge or charge state of the battery cell. The regularization term is defined as: ; in, To represent a battery cell At the moment The regularization term of ; is the regularization coefficient, which is used to adjust the range of the equilibrium strategy; It is an ideal balancing strategy, usually set based on the optimal battery model. The final balancing strategy is: ; By introducing the second regularization term The present invention ensures the smoothness of the balancing operation while avoiding excessive interference with the normal working state of the battery.

[0057] Through adaptive balancing optimization based on self-organizing neural network (SOM), the present invention breaks through the bottleneck of traditional battery balancing control: traditional balancing methods rely on rules and static algorithms, while this solution uses self-organizing neural network to make intelligent adjustments based on the health status and dynamic needs of the battery. The balancing strategy of each battery cell will change dynamically according to the real-time monitored health status and balancing needs, thereby improving the balancing efficiency and battery life.

[0058] In addition, the unsupervised learning ability of the SOM network makes this method highly adaptive and can cope with changes in the battery under different usage environments. In particular, in situations such as sudden changes in health status and load changes, it can quickly adjust the balancing strategy to avoid excessive charging and discharging of battery cells.

[0059] S104. Input the final balancing strategy, the health score of the battery cell and the balancing strength of each battery cell as a data set into a battery health prediction model based on a long short-term memory network to obtain a predicted health state of the battery cell; the battery health prediction model is used to capture the health change trend of the battery cell over a long time scale.

[0060] Specifically, the input of this step comes from the output of step S103, that is, the adaptive balancing strategy for each battery cell generated based on the SOM network These data include the balancing strategy for each battery cell. and their health status information and balanced strength .

[0061] These data can not only be used to determine the real-time health status of battery cells, but also provide important input information for future prediction models. In order to improve the accuracy of prediction management and anomaly detection, the present invention combines historical data and current status data for time series analysis. The input data is: ; Furthermore, the construction of the battery health prediction model: The prediction of battery health is based on historical data and current status. In order to improve the accuracy of the prediction, the present invention adopts a time series prediction model based on the long short-term memory network (LSTM). This model can capture the health change trend of the battery over a long time scale, and is particularly suitable for devices such as batteries that have long-term change trends. The output of the LSTM model is the battery health score. The prediction formula is:

[0062] in, To represent a battery cell At the moment predicted health status; To indicate time The input data includes balancing strategy, health status and balancing strength. Through LSTM training, the present invention can accurately capture the long-term change pattern of battery health status and predict the health trend of the battery in the future.

[0063] S105: Determine whether the error between the predicted health state of the battery cell and the actual health state of the battery cell is abnormal. If the error exceeds a threshold, an alarm is triggered and a self-recovery mechanism is entered, and adaptive adjustments are made according to actual conditions to reduce the risk of battery damage. The self-recovery mechanism adjusts the balancing strategy based on the current health state and balancing strength of the battery cell to obtain the final balancing strategy for the battery cell. .

[0064] Specifically, during the operation of the battery pack, the occurrence of abnormal conditions will have a significant impact on the health of the battery, so it is necessary to detect abnormalities in a timely manner and issue early warnings. The abnormality detection algorithm is based on the predicted health status and actual battery health status The error between them is used to determine whether an abnormality has occurred.

[0065] Define an error metric function , which is used to calculate the difference between the predicted and actual values: ; when Exceeding the set threshold When the system triggers an abnormal alarm and enters the self-recovery mechanism: if , then an alarm is triggered, indicating that the battery health status is abnormal; It is a pre-set threshold and can be adjusted dynamically according to actual conditions.

[0066] Furthermore, once an abnormality is detected, the system will automatically enable the self-recovery mechanism. The core idea of ​​self-recovery is to adjust the balancing strategy based on the current health status of the battery and the balancing requirements. According to the abnormality of the battery cell, the present invention adopts the following strategies: Mild abnormality: When the abnormality is mild, the system adjusts the battery balancing strategy to alleviate the problem and return the battery to normal working condition.

[0067] Severe abnormality: When the abnormality is serious, the system will suspend or reduce the balancing operation of the battery cell to avoid further damage and optimize the performance of the battery pack by redistributing the load.

[0068] The equilibrium strategy adjustment formula is: ; in, For battery cells At the moment The equilibrium strategy; is the change in the health error of the battery cell, indicating the change in the health status error from the last time to the current moment; The adjustment coefficient controls the adjustment range of the balancing strategy. This adjustment strategy can ensure that when an abnormality occurs, the battery operation can be adaptively adjusted according to the actual situation, thereby reducing the risk of battery damage.

[0069] By introducing predictive management, anomaly detection and self-recovery mechanisms, the present invention adds self-regulation and self-repair capabilities to the battery management system. Under the prediction of battery health change trends, the battery management system can detect potential problems in advance and correct them through the self-recovery mechanism. This automated and adaptive approach makes battery management more intelligent and can effectively avoid the problem of slow response of traditional battery management systems when facing battery anomalies.

[0070] Embodiment 2 like Figure 2 As shown, the embodiment of the present invention also provides a lithium battery pack chip balancing control system, the system comprising: A battery cell data acquisition unit 101 is used to acquire battery cell data, build a machine learning model based on the battery cell data to evaluate the battery health status, and obtain a health score of the corresponding battery cell; the machine learning model is trained using a feature vector after the battery cell data is fused; The battery cell health analysis unit 102 is used to match the health score of the battery cell with the health level, determine the health level of each battery cell, determine the balancing strength of each battery cell by optimizing the decision model based on the health score of the battery cell, and perform a first regularization operation on the balancing strength of the battery cell to obtain the final balancing control amount of the battery cell. ; The strategy training unit 103 is used to input the balancing strength of each battery cell and the health score of the battery cell into the SOM network to obtain the adaptive balancing strategy of the battery cell; the adaptive balancing strategy of the battery cell is subjected to a second regularization operation based on the final balancing control amount of the battery cell to obtain the final balancing strategy. The SOM network maps the battery cells into similar groups according to the balancing strength of each battery cell and the health score of the battery cell, and is obtained through unsupervised learning training; The fault prediction unit 104 is used to input the final balancing strategy, the health score of the battery cell and the balancing strength of each battery cell as a data set into a battery health prediction model based on a long short-term memory network to obtain a predicted health state of the battery cell; the battery health prediction model is used to capture the health change trend of the battery cell on a long time scale; The control optimization unit 105 is used to determine whether there is an abnormality based on the error between the predicted health state of the battery cell and the actual health state of the battery cell. If the error exceeds the threshold, an alarm is triggered and a self-recovery mechanism is entered, and adaptive adjustments are made according to the actual situation to reduce the risk of battery damage; the self-recovery mechanism adjusts the balancing strategy based on the current health state and balancing strength of the battery cell to obtain the final balancing strategy for the battery cell. .

[0071] The above describes specific embodiments of the present specification, and other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the accompanying drawings do not necessarily have to be performed in the specific order or sequential order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0072] The systems, devices, modules or units described in the above embodiments may be implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a computer. Specifically, the computer may be, for example, a personal computer, a laptop computer, a cellular phone, a camera phone, a smart phone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or a combination of any of these devices.

[0073] For the convenience of description, the above device is described in various units according to their functions. Of course, when implementing this specification, the functions of each unit can be implemented in the same or multiple software and / or hardware.

[0074] Those skilled in the art will appreciate that the embodiments of this specification may be provided as methods, systems, or computer program products. Therefore, the embodiments of this specification may be in the form of complete hardware embodiments, complete software embodiments, or embodiments in combination with software and hardware. Moreover, the embodiments of this specification may be in the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0075] This specification is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of this specification. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0076] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0077] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0078] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0079] The memory may include non-permanent storage in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0080] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0081] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0082] This specification may be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform specific tasks or implement specific abstract data types. This specification may also be practiced in distributed computing environments where tasks are performed by remote processing devices connected through a communication network. In a distributed computing environment, program modules may be located in local and remote computer storage media, including storage devices.

[0083] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0084] Finally, it should be noted that the lithium battery pack chip balancing control platform disclosed in the embodiment of the present invention is only a preferred embodiment of the present invention, which is only used to illustrate the technical solution of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, it should be understood by those skilled in the art that the technical solutions described in the aforementioned embodiments can still be modified, or some of the technical features therein can be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A lithium battery pack chip balancing control method, characterized in that: The method comprises the following steps: Obtain battery cell data, build a machine learning model based on the battery cell data to evaluate the battery health status, and obtain a health score of the corresponding battery cell; the machine learning model is trained using a feature vector after the battery cell data is fused; The health score of the battery cell is matched with the health level to determine the health level of each battery cell. Based on the health score of the battery cell, the balancing strength of each battery cell is determined by optimizing the decision model. The balancing strength of the battery cell is subjected to the first regularization operation to obtain the final balancing control amount of the battery cell. ; The balancing strength of each battery cell and the health score of the battery cell are input into the SOM network to obtain the adaptive balancing strategy of the battery cell; the adaptive balancing strategy of the battery cell is subjected to a second regularization operation based on the final balancing control amount of the battery cell to obtain the final balancing strategy The SOM network maps the battery cells into similar groups according to the balancing strength of each battery cell and the health score of the battery cell, and is obtained through unsupervised learning training; The final balancing strategy, the health score of the battery cell and the balancing strength of each battery cell are input as data sets into a battery health prediction model based on a long short-term memory network to obtain a predicted health state of the battery cell; the battery health prediction model is used to capture the health change trend of the battery cell over a long time scale; Based on the error between the predicted health status of the battery cell and the actual health status of the battery cell, it is judged whether it is abnormal. If the error exceeds the threshold, an alarm is triggered and the self-recovery mechanism is entered, and adaptive adjustments are made according to the actual situation to reduce the risk of battery damage; the self-recovery mechanism adjusts the balancing strategy based on the current health status and balancing strength of the battery cell to obtain the final balancing strategy for the battery cell. .

2. A lithium battery pack chip balancing control method according to claim 1, characterized in that: The battery cell data includes voltage, current, temperature and internal resistance; When acquiring battery cell data, preprocessing is also performed: The weighted moving average method is used to smooth the data of each battery cell to obtain the smoothed data; Performing standardization on the smoothed data to obtain standardized data; The standardized data are fused and analyzed and combined into a feature vector.

3. A lithium battery pack chip balancing control method according to claim 2, characterized in that: The machine learning model is trained based on the feature vector to obtain a health score of the battery cell; the health score of the battery cell ranges from [0, 1], where 0 indicates that the battery health state is extremely poor and 1 indicates that the battery is in the best state.

4. A lithium battery pack chip balancing control method according to claim 1, characterized in that: The health levels include health level 1, health level 2, and health level 3; Among them, the health level 1 indicates that the health of the battery cell is the worst and balancing is required as the top priority; the health level 2 indicates that the health of the battery cell is medium and regular balancing is performed; the health level 3 indicates that the health of the battery cell is the best and the balancing requirement is the least; The optimization decision model determines the balancing strength of each battery cell and is expressed as: ； in, Indicates battery cells At the moment Balanced strength; and is the weight parameter, , flexibility in controlling the equalization intensity; For battery cells At the moment health score; The first regularization term , which is used to control excessive intervention in the equilibrium process, is expressed as: ； in, It is the first regularization term in the balancing process, which is used to prevent over-balancing and ensure that the health status of the battery is not excessively disturbed; is the regularization coefficient, which controls the degree of intervention and is usually set between between.

5. A lithium battery pack chip balancing control method according to claim 1, characterized in that: The training goal of the SOM network is to minimize the error between neurons, so that each battery cell is mapped to the neuron closest to its health state and balance strength. During the training process, the weight of each neuron will be updated according to the input information to form the corresponding battery cell characteristics; The training is expressed as: ； in, is the updated weight; To represent neurons At the moment The weight reflects the balance demand and health status of the battery cells; is the learning rate, which controls the magnitude of each update; is the neighborhood function, which indicates the influence range of each neuron; The input data includes the battery health score and equalization strength.

6. A lithium battery pack chip balancing control method according to claim 5, characterized in that: The trained SOM network generates an adaptive balancing strategy for each battery based on the health status and balancing requirements of the battery cells. , expressed as: ； in, To represent a battery cell At the moment Adaptive balancing strategy; is the neuron weight in the SOM network, reflecting the characteristics of the battery cell; Provide health scores for battery cells; It is the balanced strength of battery cells.

7. A lithium battery pack chip balancing control method according to claim 5, characterized in that: The second regularization operation is used to control the balancing operation of each battery to be less severe, and is defined as: ； in, Indicates battery cells At the moment The second regularization term of is the regularization coefficient, which is used to adjust the range of the equilibrium strategy; Represents the ideal balancing strategy, which is usually set based on the optimal battery model.

8. A lithium battery pack chip balancing control method according to claim 1, characterized in that: The error between the predicted health state of the battery cell and the actual health state of the battery cell is measured based on an error metric function; the error metric function is expressed as: ； in, is the error measure, is the actual health status of the battery cell, The predicted health status of the battery cells.

9. A lithium battery pack chip balancing control method according to claim 1, characterized in that: The self-recovery mechanism designs a mild abnormality strategy and a severe abnormality strategy according to the abnormality degree of the battery cell; The mild abnormality strategy is to adjust the battery balancing strategy To alleviate the problem and return the battery to normal working state; the severe abnormality strategy will suspend or reduce the balancing operation of the battery cell and optimize the performance of the battery pack by redistributing the load.

10. A lithium battery chip balancing control system, characterized in that: The system comprises: A battery cell data acquisition unit, used to acquire battery cell data, build a machine learning model based on the battery cell data to evaluate the battery health status, and obtain a health score of the corresponding battery cell; the machine learning model is trained using a feature vector after the battery cell data is fused; The battery cell health analysis unit is used to match the health score of the battery cell with the health level, determine the health level of each battery cell, determine the balancing strength of each battery cell by optimizing the decision model based on the health score of the battery cell, and perform the first regularization operation on the balancing strength of the battery cell to obtain the final balancing control amount of the battery cell. ; A strategy training unit is used to input the balancing strength of each battery cell and the health score of the battery cell into the SOM network to obtain an adaptive balancing strategy for the battery cell; and perform a second regularization operation on the adaptive balancing strategy of the battery cell based on the final balancing control amount of the battery cell to obtain a final balancing strategy. The SOM network maps the battery cells into similar groups according to the balancing strength of each battery cell and the health score of the battery cell, and is obtained through unsupervised learning training; A fault prediction unit, used to input the final balancing strategy, the health score of the battery cell and the balancing strength of each battery cell as a data set into a battery health prediction model based on a long short-term memory network to obtain a predicted health state of the battery cell; the battery health prediction model is used to capture the health change trend of the battery cell on a long time scale; The control optimization unit is used to determine whether there is an abnormality based on the error between the predicted health state of the battery cell and the actual health state of the battery cell. If the error exceeds the threshold, an alarm is triggered and the self-recovery mechanism is entered, and adaptive adjustments are made according to the actual situation to reduce the risk of battery damage; the self-recovery mechanism adjusts the balancing strategy based on the current health state and balancing strength of the battery cell to obtain the final balancing strategy for the battery cell. .

Images