Series lithium battery pack power supply equalization management method

Abstract

The invention discloses a series lithium battery pack power supply equalization management method. The method comprises the following steps: acquiring an original data set of a series lithium battery pack; calculating a voltage change rate, a temperature sudden rise rate and an internal resistance fluctuation amplitude, and obtaining a dynamic feature set of battery operation; when the dynamic feature set exceeds a corresponding preset threshold value, obtaining a short-circuit risk assessment result; matching a pre-established battery pack topological structure diagram with the short-circuit risk assessment result, determining a target battery identifier needing to be isolated, generating an isolation instruction by adopting a relay control algorithm, and completing the isolation of the short-circuit battery; adjusting the charging current distribution of the updated battery pack through an equalization management algorithm to obtain equalized battery operation parameters; and if the voltage, temperature and internal resistance data of the residual single batteries are all in a normal range, determining that the system stability is recovered, and obtaining a final running state of the battery pack. The short-circuit risk of the lithium battery pack can be effectively prevented, and intelligent management and safety improvement of the battery pack are realized.

Description

Technical Field

[0001] The present invention belongs to the technical field of new energy battery management and safety control, and particularly relates to a power balancing management method for a series-connected lithium battery pack. Background Art

[0002] In the field of new energy, lithium battery pack power management is the core technology to ensure the efficient operation, safety and service life of the battery system, and plays a key role especially in applications such as electric vehicles and renewable energy energy storage. With the popularization of series-connected lithium battery packs, how to maintain the balance between single cells and the overall stability of the system has become a technical difficulty. At present, although battery balancing management methods can partially alleviate the problem of battery inconsistency, there are still significant deficiencies.

[0003] Traditional solutions usually rely on a single parameter (such as voltage) for balancing control, or use passive energy-consuming methods (such as resistor discharge) to eliminate the difference in electricity. These methods are less efficient and difficult to cope with sudden failures. Especially in extreme working conditions such as short circuits, the response delay is likely to lead to serious consequences such as thermal runaway. In series-connected lithium battery packs, the short circuit problem has become a technical bottleneck that needs to be overcome due to its strong concealment and great destructiveness. The main defects of the existing technology are reflected in the following aspects:

[0004] Single monitoring dimension: Traditional methods mostly judge based on voltage parameters, resulting in incomplete capture of short circuit precursor information and inability to achieve accurate early warning.

[0005] Insufficient short circuit feature modeling: Existing models are difficult to accurately characterize the microscopic dynamic behavior in the initial stage of short circuit, resulting in lagged fault identification.

[0006] Inefficient fault isolation mechanism: After a short circuit occurs, there is a lack of fast and reliable isolation strategies, leading to the spread of faults and threatening the safety of the entire battery pack.

[0007] In summary, the existing technology has obvious shortcomings in multi-dimensional real-time monitoring, accurate identification of short circuit precursors, and fast fault isolation, resulting in low short circuit detection accuracy and slow response speed, and it is difficult to meet the high safety requirements. Therefore, there is an urgent need to propose a power balancing management method for a series-connected lithium battery pack to improve the safety and reliability of the series-connected lithium battery pack. Summary of the Invention

[0008] To solve the above technical problems, the present invention proposes a power balancing management method for a series-connected lithium battery pack to solve the problems existing in the above-mentioned existing technology.

[0009] To achieve the above object, the present invention provides a power balancing management method for a series-connected lithium battery pack, including the following steps:

[0010] Collect multi-dimensional monitoring information of a series-connected lithium battery pack in real time, including the voltage, temperature, and internal resistance data of each single battery, and construct an original data set;

[0011] Based on the original data set, use time series analysis algorithms to calculate the voltage change rate, temperature sudden rise rate, and internal resistance fluctuation amplitude, and obtain a dynamic feature set of battery operation;

[0012] Classify and train the dynamic feature set of battery operation through the support vector machine algorithm to obtain a short-circuit precursor analysis model;

[0013] Based on the output result of the short-circuit precursor analysis model, when the voltage change rate, temperature sudden rise rate, and internal resistance fluctuation amplitude exceed the corresponding preset thresholds, it is determined that there is a short-circuit precursor in the current single battery, and a short-circuit risk assessment result is obtained;

[0014] Match the pre-established battery pack topology structure diagram with the short-circuit risk assessment result to determine the position of the single battery related to the short-circuit precursor, and obtain the target battery identifier to be isolated;

[0015] According to the target battery identifier, use the relay control algorithm to generate an isolation instruction, and complete the isolation of the short-circuit battery by disconnecting the corresponding relay connection path, and obtain the updated battery pack state;

[0016] Extract the voltage, temperature, and internal resistance data of the remaining single batteries from the updated battery pack state, and adjust the charging current distribution through the equalization management algorithm to obtain the balanced battery operation parameters;

[0017] According to the balanced battery operation parameters, if the voltage, temperature, and internal resistance data of the remaining single batteries are all within the normal range, it is determined that the system stability is restored, and the final battery pack operation state is obtained.

[0018] Optionally, the real-time collection of multi-dimensional monitoring information of the series-connected lithium battery pack, including the voltage, temperature, and internal resistance data of each single battery, and constructing the original data set includes:

[0019] By installing corresponding sensors on each single battery of the series-connected lithium battery pack, the voltage, temperature, and internal resistance data of each single battery are collected in real time;

[0020] Perform filtering, amplification, and analog-to-digital conversion processing on the voltage, temperature, and internal resistance data of each single battery to obtain multi-dimensional data;

[0021] Based on the battery equivalent circuit model, perform time series alignment and outlier rejection on the multi-dimensional data to form an original data set including the three-dimensional coupling relationship of voltage-temperature-internal resistance.

[0022] Optionally, based on the original data set, a time series analysis algorithm is used to calculate the voltage change rate, the temperature sudden rise rate, and the internal resistance fluctuation amplitude, and a dynamic feature set of battery operation is obtained, including:

[0023] Based on the original data set, the voltage change rate is calculated by first-order difference, the temperature sudden rise rate is obtained by differentiating the smoothed temperature curve using Savitzky-Golay filtering, and the internal resistance fluctuation amplitude is quantified by combining the sliding standard deviation;

[0024] Meanwhile, the dynamic time warping algorithm is introduced to align the characteristic time series differences of different battery cells, and finally a dynamic feature set of battery operation is obtained.

[0025] Optionally, based on the output result of the short-circuit precursor analysis model, when the voltage change rate, the temperature sudden rise rate, and the internal resistance fluctuation amplitude exceed the corresponding preset thresholds, it is determined that there is a short-circuit precursor in the current single battery, and a short-circuit risk assessment result is obtained, including:

[0026] Obtain the classification boundary from the short-circuit precursor analysis model and determine the separation threshold between abnormal data and normal data;

[0027] For the separation threshold, a sliding window is used to segment the voltage change rate, the temperature sudden rise rate, and the internal resistance fluctuation amplitude to obtain a segmented feature sequence;

[0028] According to the segmented feature sequence, calculate the statistical values of the features within each segment to obtain an abnormal feature subset;

[0029] If there is a feature segment in the abnormal feature subset whose statistical value exceeds the preset threshold, it is determined as an abnormal state through logical judgment to obtain an abnormal state set;

[0030] Through the abnormal state set, the K-means clustering algorithm is used to classify the abnormal data to obtain an abnormal category set;

[0031] According to the mapping relationship between the abnormal category set and the operating state, judge the state labels corresponding to each abnormal category to obtain a state classification result;

[0032] For the state classification result, combined with the operating data of the single battery, determine the short-circuit precursor risk assessment result.

[0033] Optionally, match the pre-established battery pack topology structure diagram with the short-circuit risk assessment result to determine the position of the single battery related to the short-circuit precursor, and obtain the target battery identifier to be isolated, including:

[0034] Through the pre-established battery pack topology structure diagram, obtain the position of the single battery corresponding to the short-circuit risk and determine the initial target battery identifier;

[0035] For the initial target battery identification, a preset threshold is used to determine the isolation requirement, and a battery set to be isolated is obtained;

[0036] From the battery set to be isolated, obtain the adjacency relationship of the single battery in the topological structure and determine the range of the affected battery group;

[0037] According to the scope of the affected battery pack, the propagation path of the short circuit precursor is analyzed through the structure diagram to obtain the risk diffusion trend;

[0038] In view of the risk diffusion trend, if there is a battery segment in the propagation path whose location exceeds the preset threshold, it is determined as a high-risk battery subset through logical judgment;

[0039] From the high-risk battery subset, the K-means clustering algorithm is used to classify the risk judgment results and obtain the abnormal battery category;

[0040] According to the abnormal battery category, the isolation requirements are updated through the topology diagram to determine the final target battery identification set.

[0041] Optionally, the step of generating an isolation instruction by using a relay control algorithm according to the target battery identifier, isolating the short-circuited battery by disconnecting the connection path of the corresponding relay, and obtaining an updated battery pack state includes:

[0042] Through the target battery identification, the topological data of the corresponding relay path is obtained to determine the priority order of the disconnection operation;

[0043] According to the priority order, a relay control algorithm is used to generate an isolation instruction sequence to obtain an instruction set to be executed;

[0044] For the instruction set, the disconnection condition of the relay path is determined by a preset threshold value, and if the load of the relay path exceeds the threshold value, a disconnection signal is generated;

[0045] From the disconnection signal, obtain the range of the affected connection paths and determine the isolation status of the short-circuited battery;

[0046] According to the isolation status, the K-means clustering algorithm is used to classify the battery pack status and obtain the distribution characteristics of abnormal batteries;

[0047] Based on the distribution characteristics, the topological record of the battery pack status is updated to determine the isolation completion degree of the target battery;

[0048] From the isolation completion degree, the updated battery pack operating parameters are obtained, and then the updated battery pack status is obtained.

[0049] Optionally, extract the voltage, temperature, and internal resistance data of the remaining single cells from the updated battery pack status, and adjust the charging current distribution through an equalization management algorithm to obtain the equalized battery operating parameters, including:

[0050] Obtain the voltage data, temperature data, and internal resistance data of the single cells from the battery pack status, calculate the adjustment value of the charging current using an equalization management algorithm, and obtain a preliminary current distribution plan;

[0051] For the preliminary current distribution plan, determine whether the voltage data of the single cells exceeds the range through a preset threshold. If it exceeds, adjust the charging current to obtain a corrected current distribution result;

[0052] According to the corrected current distribution result, obtain the change trend of the temperature data of the single cells, analyze the distribution characteristics of the temperature data using statistical methods, and determine the temperature equalization state;

[0053] Through the temperature equalization state, obtain the fluctuation range of the internal resistance data, determine whether the internal resistance data meets the requirements of equalization management, and obtain the operating parameters after internal resistance adjustment;

[0054] Extract the actual distribution value of the charging current from the operating parameters after internal resistance adjustment, and optimize the current distribution using an equalization management algorithm to obtain the equalized battery operating parameters.

[0055] Optionally, according to the equalized battery operating parameters, if the voltage, temperature, and internal resistance data of the remaining single cells are all within the normal range, determine that the system stability is restored, and obtain the final operating state of the battery pack, including:

[0056] Obtain the voltage difference data of the single cells from the equalized battery operating parameters, determine whether the voltage difference meets the range requirements through a preset threshold, and obtain a preliminary stability evaluation result;

[0057] According to the preliminary stability evaluation result, obtain the distribution of the temperature data, analyze the change characteristics of the temperature data using statistical methods, and determine the equalization state of the temperature distribution;

[0058] Through the equalization state of the temperature distribution, obtain the distribution of the internal resistance data, determine whether the internal resistance data is within the preset range, and obtain the adjustment basis for the internal resistance distribution;

[0059] For the adjustment basis of the internal resistance distribution, extract the real-time data of the single cells from the operating parameters to obtain an optimized parameter set;

[0060] According to the optimized parameter set, obtain the overall change trend of the battery pack status, and determine whether the system stability is restored through logical judgment to obtain a stability confirmation result;

[0061] Extract the operation conclusion from the stability confirmation result, update the recorded data of the battery pack status, and determine the balanced state of the system operation;

[0062] Based on the recorded data in the balanced state, obtain the long-term operation trend of the single battery, judge whether the battery pack status remains stable, and obtain the final operation status of the battery pack.

[0063] The present invention also provides a computer device, including a memory, a processor, and a computer program stored on the memory, and the processor executes the computer program to implement the steps of the method.

[0064] The present invention also provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the method are implemented.

[0065] Compared with the prior art, the present invention has the following advantages and technical effects:

[0066] The present invention discloses a method for balanced management of a series-connected lithium battery pack power supply. This method collects multi-dimensional monitoring information of the battery pack in real time, extracts dynamic features and constructs a short-circuit precursor analysis model. According to the output result of the model, the short-circuit risk is evaluated, and the risk battery is located in the battery pack topology diagram. Subsequently, the present invention quickly isolates the short-circuited battery through a relay control algorithm and adjusts the charging current distribution of the remaining batteries by using an equalization management algorithm. Finally, the present invention judges whether the system stability is restored to ensure the safe operation of the battery pack. This method can effectively prevent the short-circuit risk of the lithium battery pack, realize the intelligent management and safety improvement of the battery pack, extend the service life of the battery pack, and improve the energy utilization efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] The drawings constituting a part of this application are used to provide a further understanding of this application. The schematic embodiments of this application and their descriptions are used to explain this application and do not constitute an improper limitation to this application. In the drawings:

[0068] Figure 1 is a schematic flow chart of the method according to an embodiment of the present invention;

[0069] Figure 2 is a schematic flow chart of constructing a short-circuit precursor analysis model according to an embodiment of the present invention;

[0070] Figure 3 is a schematic flow chart of updating the operation status of the battery pack according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0071] It should be noted that, without conflict, the embodiments in this application and the features in the embodiments can be combined with each other. The following will refer to the drawings and combine the embodiments to detail this application.

[0072] It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. And although the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in a different order than here.

[0073] Embodiment 1

[0074] As Figure 1 shown, in this embodiment, a series lithium battery pack power equalization management method is provided, including the following steps:

[0075] Real-time collect multi-dimensional monitoring information of the series lithium battery pack, including the voltage, temperature, and internal resistance data of each single battery, and construct an original data set;

[0076] Based on the original data set, use time series analysis algorithms to calculate the voltage change rate, temperature sudden rise rate, and internal resistance fluctuation amplitude, and obtain a dynamic feature set of battery operation;

[0077] Classify and train the dynamic feature set of battery operation through the support vector machine algorithm to obtain a short-circuit precursor analysis model;

[0078] Based on the output result of the short-circuit precursor analysis model, when the voltage change rate, temperature sudden rise rate, and internal resistance fluctuation amplitude exceed the corresponding preset thresholds, it is determined that there is a short-circuit precursor in the current single battery, and a short-circuit risk assessment result is obtained;

[0079] Match the pre-established battery pack topology structure diagram with the short-circuit risk assessment result to determine the position of the single battery related to the short-circuit precursor, and obtain the target battery identifier to be isolated;

[0080] According to the target battery identifier, use the relay control algorithm to generate an isolation instruction, and complete the isolation of the short-circuit battery by disconnecting the connection path of the corresponding relay, and obtain the updated battery pack state;

[0081] Extract the voltage, temperature, and internal resistance data of the remaining single batteries from the updated battery pack state, and adjust the charging current distribution through the equalization management algorithm to obtain the equalized battery operation parameters;

[0082] According to the equalized battery operation parameters, if the voltage, temperature, and internal resistance data of the remaining single batteries are all within the normal range, it is determined that the system stability is restored, and the final battery pack operation state is obtained.

[0083] As a specific implementation manner, it specifically includes the following steps:

[0084] S101. Real - time collect multi - dimensional monitoring information of a series - connected lithium - battery pack through sensors, including the voltage, temperature, and internal resistance data of each single cell, and construct an original data set reflecting the battery state.

[0085] By installing corresponding sensors on each single cell of the series - connected lithium - battery pack, multi - dimensional monitoring information such as the voltage, temperature, and internal resistance of each single cell can be collected in real time. After being filtered, amplified, and subjected to analog - to - digital conversion by a signal conditioning circuit, the time - stamp - synchronized multi - dimensional data is transmitted to the host computer by an embedded micro - controller (such as STM32) through the CAN bus or a DAQs system. Based on a battery equivalent circuit model, such as the Thevenin model, the original data is aligned in time series and outliers are removed to form an original data set containing the three - dimensional coupling relationship of voltage - temperature - internal resistance.

[0086] Implementably, by installing sensors on each single cell of the series - connected lithium - battery pack, multi - dimensional information such as voltage, temperature, and internal resistance can be collected in real time. For example, in a battery pack composed of 10 single cells, a voltage sensor and a temperature sensor are installed on each cell. The voltage acquisition range is set between 2.5V and 4.2V, and the temperature range is from - 20°C to 60°C. The internal resistance is indirectly obtained by periodically applying a small - amplitude alternating signal and measuring the response, and the typical value may be between 20mΩ and 50mΩ. This multi - dimensional monitoring can comprehensively reflect the battery health state and avoid the performance degradation of the entire battery pack caused by single - point failure.

[0087] In a possible implementation, the signal conditioning circuit processes the collected original signal. Exemplarily, the voltage signal passes through a low - pass filter to remove high - frequency noise, and then is amplified to the range of 0V to 3.3V by an operational amplifier to adapt to the input requirements of the subsequent analog - to - digital converter. The temperature signal is converted into a voltage value through a thermistor bridge circuit and then amplified. This processing ensures the accuracy and stability of the data and provides a reliable basis for subsequent analysis.

[0088] Furthermore, after analog - to - digital conversion, an embedded micro - controller such as STM32 transmits data through the CAN bus. It can be understood that the CAN bus is widely used due to its high reliability and real - time performance. For example, 100 groups of data are transmitted per second, and each group of data includes voltage, temperature, internal resistance, and time - stamp, and the time - stamp accuracy reaches 1ms. This synchronous transmission avoids data misalignment and ensures that the information received by the host computer is highly consistent in time series, thus supporting accurate state assessment.

[0089] It should be noted that when processing data based on the Thevenin model, time series alignment is performed first. For example, if the voltage of a certain battery cell is 3.8V and the temperature is 35°C at the 5th second, but the internal resistance data appears at the 5.2nd second due to acquisition delay, it is aligned to the 5th second through interpolation. Outlier removal can be achieved by setting thresholds. For example, data with a voltage exceeding 4.3V or a temperature higher than 70°C is regarded as abnormal and removed. This method improves the reliability of the dataset.

[0090] In one embodiment, the formed three-dimensional coupled dataset of voltage-temperature-internal resistance can be used for battery state analysis. For example, at a low temperature of 20°C, the internal resistance may rise to 45mΩ and the voltage drops to 3.6V, indicating battery capacity attenuation; while at a high temperature of 50°C, the internal resistance drops to 25mΩ and the voltage remains at 4.0V, but it may indicate a risk of overheating. The analysis of this three-dimensional relationship helps predict battery life and optimize charge and discharge strategies.

[0091] Preferably, through the above solution, the host computer can display the state of each battery cell in real time and generate a trend graph. For example, after continuous operation for 100 hours, the internal resistance of a certain battery cell increases from 30mΩ to 40mΩ and the temperature rises by 5°C, indicating a need for maintenance. This visualization and prediction ability significantly improves the safety and usage efficiency of the battery pack.

[0092] In one embodiment, the integration of the sensor and the microcontroller can also support dynamic adjustment. For example, when it is detected that the temperature of a certain battery cell exceeds 45°C, the system automatically reduces the charging current from 2A to 1A, thereby reducing heat accumulation. This adaptive management extends battery life and reduces the failure rate.

[0093] S102. Extract dynamic features from the original dataset, and use time series analysis methods to calculate the voltage change rate, temperature sudden rise rate, and internal resistance fluctuation amplitude to obtain the dynamic feature set of battery operation.

[0094] Based on the original dataset, the voltage, temperature, and internal resistance data are segmented using a sliding time window (such as a 5-second window); the voltage change rate (dV / dt) is calculated through first-order difference to reflect the charge and discharge dynamic characteristics, the temperature sudden rise rate (dT / dt) is obtained by differentiating the smoothed temperature curve using the Savitzky-Golay filter to capture the precursors of thermal runaway, and the internal resistance fluctuation amplitude (ΔR) is quantified by combining the sliding standard deviation (rolling window of 30 seconds) to characterize the degree of battery aging; at the same time, the dynamic time warping (DTW) algorithm is introduced to align the characteristic time series differences of different battery cells, and finally a dynamic feature set containing time-domain gradient features and statistical features is constructed, providing a multi-dimensional feature vector reflecting transient characteristics and trend evolution for subsequent battery state evaluation. This process transforms static monitoring data into dynamic behavior indicators by integrating signal processing and time series analysis, effectively enhancing the sensitivity to abnormal battery conditions.

[0095] Implementable, the sliding time window is a commonly used method for processing time series data, which segments the data for analysis by setting a fixed-length time period. For example, within a 5-second window, voltage, temperature, and internal resistance data can be divided into multiple consecutive segments, each containing the sampling points within 5 seconds. This method can capture the short-term change trends.

[0096] It can be understood that for a system with a sampling frequency of 10 Hz, there will be 50 data points within a 5-second window. The voltage may slowly rise from 3.7 V to 3.8 V, the temperature may slightly increase from 30 °C to 31 °C, and the internal resistance remains near 35 mΩ. Through segmented processing, the system can observe the battery behavior with finer granularity. When calculating the voltage change rate by first-order difference, the purpose is to reflect the dynamic characteristics of the battery during charge and discharge.

[0097] Exemplarily, for the voltage data within a window, assuming the adjacent sampling points are 3.75 V and 3.76 V respectively, and the time interval is 0.1 second, the voltage change rate is approximately 0.1 V / s. This change rate may be small at the initial stage of charging and may increase significantly when approaching full charge, indicating a change in the charging state.

[0098] Specifically, if the voltage of a certain battery rises from 3.6 V to 3.9 V within 5 seconds, the change rate can reach 0.06 V / s, indicating that it is in the fast charging stage. This analysis helps to judge the real-time working state of the battery. The Savitzky-Golay filter is used to smooth the temperature curve and calculate the temperature sudden rise rate to identify the risk of thermal runaway in advance.

[0099] In a possible implementation, assuming that the temperature data has slight fluctuations, such as 30.1 °C, 30.3 °C, 30.2 °C, the smoothed value is 30.2 °C after filtering, and then the derivative is calculated to obtain the sudden rise rate. If the temperature rapidly rises from 32 °C to 35 °C within a certain period of time, and the rate reaches 0.6 °C / s, it may indicate insufficient heat dissipation or abnormal heating. This smoothing process avoids noise interference and improves the sensitivity of anomaly detection.

[0100] The sliding standard deviation is used to quantify the amplitude of internal resistance fluctuations and characterize the degree of battery aging. For example, within a 30-second window, the internal resistance value fluctuates between 33 mΩ and 37 mΩ, and the standard deviation is about 1.5 mΩ, indicating good internal resistance stability. However, if the fluctuation range expands to 30 mΩ to 45 mΩ and the standard deviation increases to 5 mΩ, it may reflect the deterioration of electrode materials or poor contact.

[0101] Preferably, this statistical feature can intuitively reflect the trend of battery performance degradation over time.

[0102] The dynamic time warping algorithm is used to align the characteristic time series differences of different battery cells. It should be noted that due to the possible slight deviation in the sampling time of each battery cell, for example, the voltage peak of one battery cell appears at the 10th second, while that of another appears at the 10.2nd second, DTW achieves alignment by stretching or compressing the time axis. For example, in a 10-cell battery pack, the internal resistance change trends of the 1st and 5th cells are similar but there is a time offset of 0.3 seconds. DTW can match the characteristic sequences of the two and unify them to the same time reference. This alignment ensures the consistency of the data of multiple battery cells.

[0103] In one embodiment, the constructed dynamic feature set fuses time-domain gradient features and statistical features. For example, the feature vector of a certain battery cell may include a voltage change rate of 0.05 V / s, a temperature sudden rise rate of 0.2 °C / s, an internal resistance fluctuation amplitude of 2 mΩ, etc. These features not only reflect transient characteristics but also can reveal the evolution of long-term trends, providing a multi-dimensional perspective for subsequent state assessment.

[0104] It can be understood that this feature set enhances the sensitivity to abnormal operating conditions, such as quickly identifying situations of sudden voltage drop or abnormal temperature rise, thereby improving the accuracy of battery management.

[0105] Specifically, this method combines signal processing and time series analysis to convert static data into dynamic indicators. For example, during the charging process, if the voltage change rate of a certain battery cell suddenly drops from 0.04 V / s to 0.01 V / s and at the same time the internal resistance fluctuation increases, the system can promptly judge that there may be capacity attenuation. The construction of this multi-dimensional feature vector lays the foundation for the comprehensive assessment of the battery state.

[0106] S103. Classify and train the voltage change rate, temperature sudden rise rate, and internal resistance fluctuation amplitude through the support vector machine algorithm to obtain a short-circuit precursor analysis model.

[0107] As Figure 2 shown, based on the dynamic feature set, using the support vector machine (SVM) algorithm, with the voltage change rate (dV / dt), temperature sudden rise rate (dT / dt), and internal resistance fluctuation amplitude (ΔR) as input feature vectors, the non-linearly separable high-dimensional features are mapped to the renewable Hilbert space through a kernel function (such as the RBF kernel), and the classification hyperplane is optimized using the principle of structural risk minimization; in the training stage, historical data with known labels (normal / short-circuit precursor) is used, and the optimal penalty coefficient C and kernel parameter γ are determined by combining grid search and cross-validation, so that the model maximizes the classification margin while reducing the risk of overfitting. Finally, a short-circuit precursor analysis model that can identify sudden voltage drop, abnormal temperature rise, and internal resistance mutation is obtained. This model realizes the precise early warning of battery early faults by fusing multi-dimensional dynamic features and the small-sample classification advantage of SVM.

[0108] Feasible. When constructing the dynamic feature set, the voltage change rate, the temperature sudden rise rate, and the internal resistance fluctuation amplitude can be used as input feature vectors and extracted from the real-time data of battery operation. For example, in a charging scenario, assume that the voltage of a certain battery rises from 3.6V to 3.8V in 4 seconds, the temperature rises from 31°C to 33°C, and the internal resistance slightly changes between 34mΩ and 36mΩ. Through time series segmentation, the voltage change rate can be initially estimated as 0.05V / s, the temperature sudden rise rate as 0.5°C / s, and the internal resistance fluctuation amplitude as 1mΩ. This feature extraction method provides multi-dimensional input for subsequent SVM classification.

[0109] In a possible implementation, the SVM maps these non-linear features to a high-dimensional space through the RBF kernel function. The core of the RBF kernel lies in measuring the distance difference between samples. For example, for two sets of feature vectors, one set is 0.03V / s, 0.2°C / s, 1mΩ under normal conditions, and the other set is 0.1V / s, 1°C / s, 3mΩ as a precursor to short circuit. The kernel function will amplify the difference between the two, facilitating the division by the classification hyperplane. This mapping enhances the sensitivity of the model to abnormal features.

[0110] It should be noted that the principle of structural risk minimization is optimized in the SVM by balancing the classification margin and the misclassification penalty.

[0111] Specifically, the penalty coefficient C controls the tolerance of the model to misclassifications. For example, if C is set to 1, the model tends to pursue a larger margin; if C is increased to 10, it pays more attention to reducing misclassifications. In a battery data set, assuming that normal samples account for 80% and samples with precursors to short circuit account for 20%, a higher C value can better focus on rare abnormal samples and improve the early warning ability.

[0112] Preferably, grid search and cross-validation are used for parameter tuning. For example, in a training set containing 1000 groups of historical data, the trial range of C is set to 0.1 to 10, and γ is set to 0.01 to 1. The accuracy of each set of parameters is evaluated through 5-fold cross-validation.

[0113] In one embodiment, when C is 2 and γ is 0.1, the classification accuracy of the model on the validation set reaches 95%, indicating that the parameter combination effectively captures the laws between features.

[0114] Specifically, the training phase relies on data with known labels. For example, under normal conditions, the voltage change rate is mostly between 0.02V / s and 0.05V / s, the temperature sudden rise rate is lower than 0.3℃ / s, while the precursor of short circuit may be manifested as the voltage change rate suddenly increasing to 0.08V / s and the temperature rate exceeding 0.8℃ / s. Through these labels, SVM learns to distinguish the boundary between normal and abnormal. In one case, the voltage change rate of a certain battery suddenly drops from 0.04V / s to 0.01V / s at the end of charging, and at the same time the temperature sudden rise rate reaches 1.2℃ / s, and the model successfully marks it as a short circuit risk.

[0115] It can be understood that the small sample classification advantage of SVM is particularly prominent in battery fault warning. For example, a model trained with only 200 sets of data can identify 90% of the abnormal temperature rise samples in the test. This high efficiency stems from the dependence of SVM on support vectors rather than the total amount of data.

[0116] In one embodiment, the feature vector fuses the information of voltage sudden drop, abnormal temperature rise and internal resistance mutation. The model can issue a warning 5 seconds before the abnormality occurs, winning time for system intervention. For example, in the monitoring of a 10-cell battery pack, the temperature sudden rise rate of the 3rd battery quickly rises from 0.2℃ / s to 1℃ / s, and the internal resistance fluctuation increases from 2mΩ to 5mΩ. Through feature vector analysis, the SVM model quickly classifies it as a precursor of short circuit. The combination of this multi-dimensional feature fusion and SVM ensures the accuracy and reliability of early fault identification.

[0117] S104. For the output result of the short circuit precursor analysis model, if it is determined that the voltage change rate exceeds the preset threshold or the temperature sudden rise rate exceeds the preset threshold or the internal resistance fluctuation amplitude is abnormal, then it is determined that the current single cell has a short circuit precursor, and a short circuit risk assessment result is obtained.

[0118] Obtain the classification boundary from the short circuit precursor analysis model and determine the separation threshold between abnormal data and normal data. For the separation threshold, use the sliding window technique to segment the voltage change rate, temperature sudden rise rate, and internal resistance fluctuation amplitude to obtain a segmented feature sequence. According to the segmented feature sequence, calculate the statistical values of the features within each segment to obtain an abnormal feature subset. If the statistical value of a certain segment of features in the abnormal feature subset exceeds the preset threshold, then through logical judgment, it is determined that the data in this segment is in an abnormal state, and an abnormal state set is obtained. Through the abnormal state set, use the K-means clustering algorithm to classify the abnormal data to obtain an abnormal category set. According to the mapping relationship between the abnormal category set and the operating state, judge the state labels corresponding to each abnormal category to obtain a state classification result. For the state classification result, combine the operating data of the single cell to determine the short circuit precursor risk assessment result.

[0119] Implementable. When obtaining the classification boundary from the short - circuit precursor analysis model, the separation threshold between abnormal data and normal data can be determined by analyzing the distribution of support vectors during the training process. Exemplarily, in a battery monitoring scenario, assuming that under normal conditions, the voltage change rate is mostly between 0.02V / s and 0.05V / s, while in an abnormal state, it may exceed 0.08V / s, the separation threshold can be initially set to 0.06V / s. This threshold setting is based on the distribution characteristics of historical data to ensure discrimination.

[0120] In a possible implementation, the sliding window technique is used to segment the real - time data stream. For example, with a window length of 2 seconds and a step size of 1 second, the voltage change rate, temperature sudden - rise rate, and internal resistance fluctuation amplitude are segmented. Suppose a data record for a certain segment shows that the voltage rises from 3.7V to 3.75V in 2 seconds, the temperature rises from 32°C to 33°C, and the internal resistance changes from 35mΩ to 36mΩ. Then the characteristics of this segment are 0.025V / s, 0.5°C / s, and 1mΩ respectively. This segmentation method is convenient for capturing dynamic changes.

[0121] It should be noted that the statistical value calculation of the segmented feature sequence can include the mean, maximum value, variance, etc. Specifically, within a 10 - second monitoring window, if the maximum value of the temperature sudden - rise rate in a certain segment reaches 1°C / s, exceeding the normal range by 0.3°C / s, it is marked as an abnormal feature subset. This statistical analysis helps to screen key abnormal points.

[0122] Preferably, logical judgment is used to determine the abnormal state. For example, if in a certain segment of feature statistics, the voltage change rate reaches 0.07V / s, the temperature sudden - rise rate exceeds 0.8°C / s, and at the same time, the internal resistance fluctuation is 3mΩ, exceeding the preset thresholds of 0.06V / s, 0.5°C / s, and 2mΩ, then this segment is classified into the abnormal state set. This multi - condition judgment improves the reliability of abnormal recognition.

[0123] In one embodiment, the K - means clustering algorithm is used to classify the abnormal state set. Assume that the abnormal data is divided into two categories, one dominated by voltage mutation and the other dominated by temperature rise anomaly. For example, an abnormal data point is 0.1V / s, 0.2°C / s, 1mΩ, and another data point is 0.03V / s, 1.2°C / s, 2mΩ. After clustering, the abnormal patterns can be clearly distinguished. This classification helps to refine the abnormal features.

[0124] It can be understood that the mapping relationship between the abnormal category and the operating state needs to be predefined. For example, the voltage mutation category may correspond to the risk of over - voltage during charging, and the temperature rise anomaly category may indicate the precursor of a short - circuit. In a case, the temperature sudden - rise rate of a certain battery reaches 1.5°C / s and is classified into the short - circuit risk category, which is mapped to the high - risk state label. This mapping provides a basis for subsequent decision - making.

[0125] Specifically, when conducting risk assessment by combining the operation data of individual cells, the frequency and duration of anomalies can be analyzed. For example, if a certain cell has an anomaly where the temperature sudden increase rate exceeds 1°C / s three times within 5 minutes and each time lasts for more than 3 seconds, the risk assessment result may be "high risk". This comprehensive analysis enhances the practicality of early warning.

[0126] In one embodiment, the effectiveness of risk assessment is verified from multiple aspects. For example, by combining historical data for verification, if a certain cell has failed due to an internal resistance fluctuation exceeding 4 mΩ, and the internal resistance of a certain section reaches 5 mΩ in the current monitoring, it is inferred that the risks are consistent; another example is that by comparing with other battery packs, if the abnormal feature subset is unique within the group, its abnormality is further confirmed. This multi-dimensional verification ensures the rigor of the assessment.

[0127] S105. After obtaining the short-circuit risk assessment result, through the pre-established battery pack topology structure diagram, determine the positions of the individual cells related to the short-circuit precursor, and obtain the target cell identifiers to be isolated.

[0128] Through the pre-established battery pack topology structure diagram, obtain the positions of the individual cells corresponding to the short-circuit risk, and determine the target cell identifiers. For the target cell identifiers, use a preset threshold to judge the isolation requirement, and obtain the set of cells to be isolated. From the set of cells to be isolated, obtain the adjacency relationship of the individual cells in the topology structure, and determine the range of the affected battery packs. According to the range of the affected battery packs, analyze the propagation path associated with the precursor through the structure diagram to obtain the risk diffusion trend. For the risk diffusion trend, if the position of a certain section of cells in the propagation path exceeds the preset threshold, determine the high-risk battery subset through logical judgment. From the high-risk battery subset, use the K-means clustering algorithm to classify the risk judgment results and obtain the abnormal battery categories. According to the abnormal battery categories, update the isolation requirement through the topology structure diagram to determine the final set of target cell identifiers.

[0129] Implementably, obtaining the positions of the individual cells corresponding to the short-circuit risk through the pre-established battery pack topology structure diagram is the basis of risk analysis. For example, in a battery pack composed of 20 individual cells connected in series, assuming that the No. 5 cell is marked as a short-circuit risk point, its position in the series chain can be quickly located through the topology diagram, and the target cell identifier is "Battery 5". This positioning method relies on the topology diagram clearly marking the connection relationship of each cell, which is convenient for subsequent analysis. For the target cell identifiers, use a preset threshold to judge the isolation requirement and determine the set of cells to be isolated.

[0130] Specifically, assume that the preset thresholds are a voltage change rate of 0.06 V / s and a temperature sudden rise rate of 0.5 °C / s. If the real-time data of "Battery 5" is 0.08 V / s and 0.7 °C / s, exceeding the thresholds, it will be included in the set of batteries to be isolated. This threshold judgment uses quantitative indicators to quickly screen out abnormal batteries and ensure the timeliness of isolation. From the set of batteries to be isolated, obtain the adjacency relationships of individual batteries in the topological structure to determine the scope of the affected battery group.

[0131] Exemplarily, if "Battery 5" is directly connected in series with "Battery 4" and "Battery 6", and the topology diagram shows that its abnormality may affect adjacent batteries through current or heat conduction, the affected range is initially determined as "Battery 4 - 6". This adjacency analysis helps to clarify the local impact boundary of the risk. According to the scope of the affected battery group, analyze the propagation path of the precursor association through the structure diagram to obtain the risk diffusion trend.

[0132] In a possible implementation, if the temperature sudden rise of "Battery 5" spreads to "Battery 6", and the temperature sudden rise rate of "Battery 6" rises to 0.6 °C / s, exceeding the threshold of 0.5 °C / s, the propagation path is "Battery 5 → Battery 6". This path analysis reveals the dynamic expansion direction of the risk. For the risk diffusion trend, if the position of a certain section of batteries in the propagation path exceeds the preset threshold, determine the high-risk battery subset through logical judgment.

[0133] Preferably, if in addition to the temperature anomaly of "Battery 6", the internal resistance fluctuation reaches 3 mΩ, exceeding the threshold of 2 mΩ, it will be grouped into the high-risk subset together with "Battery 5". This multi-condition judgment improves the accuracy of screening. From the high-risk battery subset, use the K-means clustering algorithm to classify the risk judgment results to obtain the abnormal battery categories. For example, "Battery 5" is mainly characterized by voltage mutation, with features of 0.08 V / s and 0.7 °C / s, and "Battery 6" is mainly characterized by temperature rise, with features of 0.03 V / s and 0.6 °C / s. After clustering, they are divided into "voltage anomaly class" and "temperature rise anomaly class". This classification refines the abnormal features and facilitates targeted processing. According to the abnormal battery categories, update the isolation requirements through the topological structure diagram to determine the final set of target battery identifiers.

[0134] It can be understood that if the "voltage anomaly class" maps to overvoltage risk and the "temperature rise anomaly class" points to short-circuit precursor, both "Battery 5" and "Battery 6" need to be isolated, and the final set is "Battery 5, Battery 6". This update combines the classification results with the topological relationship to ensure the comprehensiveness of the isolation strategy.

[0135] It should be noted that the use of the topology diagram runs through the whole process, from positioning to diffusion analysis, and then to isolation decision-making, progressing step by step. This method expands the risk from a single point to the whole through structured analysis, enhancing the reliability of battery pack management.

[0136] In one embodiment, if the adjacent batteries do not exceed the threshold, it can be used as an expansion solution to isolate only the core abnormal battery, taking into account both efficiency and safety.

[0137] S106. Generate an isolation instruction using a relay control algorithm according to the target battery identifier, disconnect the connection path of the corresponding relay, and quickly isolate the short-circuited battery to obtain an updated battery pack status.

[0138] Through the target battery identification, the topological data of the corresponding relay path is obtained to determine the priority order of the disconnection operation. According to the priority order, the relay control algorithm is used to generate an isolation instruction sequence to obtain a set of instructions to be executed. For the instruction set, the disconnection condition of the relay path is judged by the preset threshold. If the load of a certain path exceeds the threshold, a disconnection signal is generated. From the disconnection signal, the affected connection path range is obtained to determine the isolation status of the short-circuited battery. According to the isolation status, the K-means clustering algorithm is used to classify the battery pack status to obtain the distribution characteristics of the abnormal battery. According to the distribution characteristics, the topological record of the battery pack status is updated to determine the isolation completion of the target battery. From the isolation completion, the updated battery pack operating parameters are obtained to determine the system stability state.

[0139] It is feasible to obtain the topological data of the corresponding relay path through the target battery identification to determine the priority order of the disconnection operation. It is understandable that this process depends on the distribution and connection logic of the relays in the battery pack.

[0140] For example, in a battery pack consisting of 10 single cells connected in series, assuming that "Battery 3" is marked as the target battery, its corresponding relay path includes the main circuit relay R1 and the branch relay R2. The topological data shows that R1 controls the entire series chain, and R2 is only responsible for the connection between "Battery 3" and the adjacent battery. In terms of priority, R2 is superior to R1 because local disconnection can isolate the abnormality faster and reduce the impact on the system. According to the priority order, the relay control algorithm is used to generate an isolation instruction sequence to obtain a set of instructions to be executed.

[0141] Specifically, the algorithm will sort according to path importance and response speed. In one embodiment, if R2 takes 0.1 seconds to disconnect and R1 takes 0.3 seconds, the instruction sequence is "R2 first, then R1", and a set is generated such as "R2 disconnect, R1 standby". This ensures the high efficiency of isolation. For the instruction set, the disconnection condition of the relay path is determined by the preset threshold. For example, assuming that the load current threshold is 50A, if the real-time current of the R2 path reaches 60A, exceeding the threshold, a "R2 disconnect" signal is generated. This quantitative judgment improves the accuracy of the operation.

[0142] Obtain the range of affected connection paths from the disconnection signal to determine the isolation status of short - circuited batteries. Preferably, if the connection between "Battery 3" and "Battery 4" is interrupted after R2 is disconnected and the current data goes to zero, the isolation status is "isolated". This helps to quickly verify the effect.

[0143] According to the isolation status, use the K - means clustering algorithm to classify the battery pack status and obtain the abnormal battery distribution characteristics. In one possible implementation, the clustering is based on voltage and temperature data. If the voltage of "Battery 3" is 3.2V and the temperature is 50°C, while the normal batteries are around 3.8V and 30°C, the classification result shows that "Battery 3" is an isolated abnormal point. This classification clearly reveals the abnormal distribution.

[0144] Update the topological record of the battery pack status through the distribution characteristics to judge the isolation completion degree of the target battery. It should be noted that the topological record will mark that "Battery 3" has been disconnected. If the data of its neighboring batteries is stable, the completion degree can be set at 100%. This provides a basis for subsequent management.

[0145] Obtain the updated operating parameters of the battery pack from the isolation completion degree to determine the system stable state. Exemplarily, if the total voltage recovers from 36V to 35V and the average temperature drops to 32°C, the system state can be determined as "stable". This parameter analysis ensures the safety of the overall operation.

[0146] In one embodiment, if the parameters of neighboring batteries fluctuate slightly but do not exceed the threshold, only the core abnormal battery can be isolated. As an extended solution, it is both efficient and flexible. This multi - level analysis from local to global ensures the reliability of the battery pack.

[0147] S107: Extract the voltage, temperature, and internal resistance data of the remaining single batteries from the updated battery pack status, and adjust the charging current distribution through the equalization management algorithm to obtain the equalized battery operating parameters.

[0148] Obtain the voltage data, temperature data, and internal resistance data of the single batteries from the battery pack status, calculate the adjustment value of the charging current using the equalization management algorithm to obtain a preliminary current distribution plan. For the preliminary current distribution plan, judge whether the voltage data of the single batteries exceeds the range through a preset threshold. If it exceeds, adjust the charging current to obtain a corrected current distribution result. According to the corrected current distribution result, obtain the change trend of the temperature data of the single batteries, analyze the distribution characteristics of the temperature data using statistical methods to determine the temperature equalization state. Through the temperature equalization state, obtain the fluctuation range of the internal resistance data, judge whether the internal resistance data meets the requirements of equalization management to obtain the operating parameters after internal resistance adjustment. Extract the actual distribution value of the charging current from the operating parameters after internal resistance adjustment, and optimize the current distribution using the equalization management algorithm to obtain the equalized battery operating parameters.

[0149] It is feasible to obtain the voltage data, temperature data, and internal resistance data of individual cells from the battery pack state. It can be understood that these data are the basis for balancing management.

[0150] Exemplarily, in a battery pack composed of 8 series-connected individual cells, assuming the voltage data are 3.7V, 3.8V, 3.6V, etc., the temperature data are 28°C, 30°C, 32°C, etc., and the internal resistance data are 20mΩ, 22mΩ, 19mΩ, etc. When calculating the charging current adjustment value using the balancing management algorithm.

[0151] Specifically, the current to be increased or decreased can be determined by comparing the voltage differences of each battery. For example, if the target voltage average is 3.75V, the battery with a voltage of 3.6V needs to increase the charging current, and the preliminary distribution plan may be "Battery 1 increases by 0.2A, Battery 2 decreases by 0.1A".

[0152] For the preliminary current distribution plan, it is judged whether the voltage exceeds the range through a preset threshold. In a possible implementation, assume the safe voltage range is 3.5V to 4.0V. If the voltage of a certain battery reaches 4.1V, then its charging current is adjusted to decrease by 0.3A, and the corrected distribution result may be "Battery 1 is 0.5A, Battery 2 is 0.4A". This adjustment ensures that the voltage remains within the safe range.

[0153] According to the corrected current distribution result, obtain the trend of temperature data change. Preferably, the temperature change within 10 minutes after adjustment can be observed. For example, the temperature of Battery 1 rises from 28°C to 29°C, and the temperature of Battery 2 drops from 30°C to 29.5°C.

[0154] Use statistical methods to analyze the distribution characteristics. Specifically, the temperature average of 29.5°C and the standard deviation of 0.5°C can be calculated, indicating that the temperature balance state is good. This analysis helps to judge the effectiveness of thermal management.

[0155] Obtain the fluctuation range of the internal resistance data through the temperature balance state. It should be noted that the internal resistance is greatly affected by temperature. In one embodiment, if after the temperature stabilizes, the fluctuation range of the internal resistance shrinks from 19mΩ - 22mΩ to 20mΩ - 21mΩ, it is considered to meet the balancing management requirements. The adjusted operating parameters may be "Internal resistance average 20.5mΩ, fluctuation ±0.5mΩ". This stability improves the battery consistency.

[0156] Extract the actual distribution value of the charging current from the adjusted operating parameters of the internal resistance. For example, the actual current of Battery 1 is 0.45A, and the actual current of Battery 2 is 0.42A.

[0157] Optimized by the balanced management algorithm. In one possible implementation, it can be comprehensively adjusted according to the internal resistance and voltage, and finally the balanced operating parameters are obtained, such as "the current of battery 1 is 0.43A, the voltage is 3.75V, and the temperature is 29°C". This optimization improves the overall efficiency. When inferring from multiple aspects by way of example, assume that the temperature of a certain battery suddenly rises to 35°C and the voltage is still normal. Its current can be adjusted to 0.3A only, and then observe whether the internal resistance decreases. If the internal resistance returns to normal, it is the core solution; if it is still on the high side, it can be extended to reduce the current of adjacent batteries to share the heat load jointly. This multi-level method is both accurate and flexible, ensuring the long-term reliability of the battery pack.

[0158] S108. According to the balanced battery operating parameters, if it is determined that the voltage difference is less than the preset threshold and both the temperature and the internal resistance are within the normal range, then it is determined that the system stability is restored, and the final operating state of the battery pack is obtained.

[0159] As Figure 3 shown, obtain the voltage difference data of the single cells from the balanced operating parameters, judge whether the voltage difference meets the range requirements through the preset threshold, and obtain the preliminary stability evaluation result. According to the preliminary stability evaluation result, obtain the distribution of the temperature data, analyze the change characteristics of the temperature data by using statistical methods, and determine the balanced state of the temperature distribution. Through the balanced state of the temperature distribution, obtain the distribution of the internal resistance data, judge whether the internal resistance data is within the preset range, and obtain the adjustment basis for the internal resistance distribution. For the adjustment basis of the internal resistance distribution, extract the real-time data of the single cells from the operating parameters, optimize the data distribution by using the balanced algorithm, and obtain the optimized parameter set. According to the optimized parameter set, obtain the overall change trend of the battery pack state, and determine whether the system stability is restored through logical judgment to obtain the stability confirmation result. Extract the operating conclusion from the stability confirmation result, update the recorded data of the battery pack state, and determine the balanced state of the system operation. Through the recorded data in the balanced state, obtain the long-term operation trend of the single cells, and judge whether the battery pack state is continuously stable to obtain the final operating parameter adjustment plan.

[0160] Implementable, obtain the voltage difference data of the single cells from the balanced operating parameters. It can be understood that the voltage difference reflects the consistency of each single cell in the battery pack. For example, in a battery pack with 8 cells connected in series, the balanced voltages may be 3.74V, 3.76V, 3.73V, etc., and the difference range is from 0.01V to 0.03V.

[0161] Preferably, judge by the preset threshold such as 0.05V. If the difference is less than this value, it is initially considered that the stability is good. This evaluation lays the foundation for subsequent analysis.

[0162] Obtain the distribution of temperature data based on the results of preliminary stability assessment. Specifically, the temperatures of each battery can be observed, such as 28.5°C, 29°C, 28.8°C, etc. In one possible implementation, statistical methods are used to calculate the mean value of 28.8°C and the maximum deviation of 0.2°C, indicating a uniform temperature distribution. This uniformity helps reduce the risk of local overheating.

[0163] When obtaining the distribution of internal resistance data through the equilibrium state of the temperature distribution. It should be noted that the internal resistance is closely related to temperature. Exemplarily, if the temperature is stable, the internal resistance data may be 20.2 mΩ, 20.5 mΩ, 20.1 mΩ, with a fluctuation range of only 0.4 mΩ. Determine whether it is within a preset range, such as ±0.5 mΩ. If it is satisfied, the internal resistance distribution is reasonable, providing a reliable basis for adjustment.

[0164] When extracting real-time data for the adjustment basis of the internal resistance distribution. For example, the voltage of a certain battery is 3.73 V, the temperature is 29°C, and the internal resistance is 20.3 mΩ.

[0165] When optimizing using an equalization algorithm. In one embodiment, the current can be fine-tuned to 0.4 A according to the high internal resistance situation, and optimized parameters such as voltage 3.75 V and internal resistance 20.2 mΩ can be obtained. This optimization improves data consistency.

[0166] When obtaining the overall change trend of the battery pack state based on the optimized parameter set. Preferably, it can be observed that the average voltage rises from 3.74 V to 3.75 V, and the temperature fluctuation shrinks to 0.1°C. Through logical judgment, if the trend is stable, the system stability is restored. This confirmation ensures the operation reliability.

[0167] When extracting the operation conclusion from the stability confirmation result to update the record data. For example, the record shows that the voltage difference has been maintained within 0.02 V for a long time, and the average temperature is stable at 29°C. This update provides data support for long-term monitoring.

[0168] When obtaining the long-term operation trend through the record data in the equilibrium state. In one possible implementation, the voltage of a certain battery is 3.75 V and the internal resistance is 20.2 mΩ, without obvious fluctuations for 10 hours, indicating a continuously stable state.

[0169] Exemplarily, if the temperature of a certain battery suddenly rises to 31°C, its current can be reduced to 0.35 A to observe the trend; if the internal resistance then drops to 20.1 mΩ, it is the core solution. If it is still on the high side, it can be extended to adjust the current of the adjacent battery to 0.38 A to share the load. This multi-level adjustment ensures long-term stability. When considering from multiple aspects, for example, if the voltage of a certain battery drops to 3.70 V and the temperature is normal, the current can be increased to 0.45 A to observe the change in internal resistance. If the internal resistance is stable, the solution is effective; if the fluctuation increases, the temperature data analysis can be combined and adjusted to 0.42 A. This multi-directional verification enhances the flexibility and accuracy of the solution.

[0170] Embodiment 2

[0171] This embodiment also provides a computer device, including a memory, a processor, and a computer program stored on the memory. The processor executes the computer program to implement the steps of the method.

[0172] Embodiment 3

[0173] This embodiment also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the method are implemented.

[0174] The above is only the preferred specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by those skilled in the art within the technical scope disclosed in the present application should be covered by the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A method for balancing management of a series-connected lithium battery pack power supply, characterized in that, Including the following steps: Collect multi-dimensional monitoring information of the series-connected lithium battery pack in real time, including the voltage, temperature, and internal resistance data of each single battery, and construct an original data set; Based on the original data set, use time series analysis algorithms to calculate the voltage change rate, temperature sudden rise rate, and internal resistance fluctuation amplitude, and obtain the dynamic feature set of battery operation; Classify and train the dynamic feature set of battery operation through the support vector machine algorithm to obtain a short-circuit precursor analysis model; Based on the output result of the short-circuit precursor analysis model, when the voltage change rate, temperature sudden rise rate, and internal resistance fluctuation amplitude exceed the corresponding preset thresholds, it is determined that there is a short-circuit precursor in the current single battery, and a short-circuit risk assessment result is obtained; Match the pre-established battery pack topology structure diagram with the short-circuit risk assessment result to determine the position of the single battery related to the short-circuit precursor, and obtain the target battery identification to be isolated; According to the target battery identification, use the relay control algorithm to generate an isolation instruction, and complete the isolation of the short-circuit battery by disconnecting the corresponding relay connection path, and obtain the updated battery pack state; Extract the voltage, temperature, and internal resistance data of the remaining single batteries from the updated battery pack state, and adjust the charging current distribution through the equalization management algorithm to obtain the balanced battery operation parameters; According to the balanced battery operation parameters, if the voltage, temperature, and internal resistance data of the remaining single batteries are all within the normal range, it is determined that the system stability is restored, and the final battery pack operation state is obtained.

2. The method according to claim 1, wherein The real-time collection of multi-dimensional monitoring information of the series-connected lithium battery pack, including the voltage, temperature, and internal resistance data of each single battery, and the construction of the original data set, includes: By installing corresponding sensors on each single battery of the series-connected lithium battery pack, the voltage, temperature, and internal resistance data of each single battery are collected in real time; Perform filtering, amplification, and analog-to-digital conversion processing on the voltage, temperature, and internal resistance data of each single battery to obtain multi-dimensional data; Based on the battery equivalent circuit model, perform time series alignment and outlier rejection on the multi-dimensional data to form an original data set containing the three-dimensional coupling relationship of voltage-temperature-internal resistance.

3. The method according to claim 1, wherein The calculation of the voltage change rate, temperature sudden rise rate, and internal resistance fluctuation amplitude based on the original data set by using time series analysis algorithms to obtain the dynamic feature set of battery operation includes: Based on the original data set, calculate the voltage change rate through first-order difference, obtain the temperature sudden rise rate by differentiating the smoothed temperature curve using Savitzky-Golay filtering, and quantify the internal resistance fluctuation amplitude by combining the sliding standard deviation; At the same time, introduce the dynamic time warping algorithm to align the characteristic time series differences of different battery monomers, and finally obtain the dynamic feature set of battery operation.

4. The method according to claim 1, wherein Based on the output result of the short-circuit precursor analysis model, when the voltage change rate, temperature sudden rise rate, and internal resistance fluctuation amplitude exceed the corresponding preset thresholds, it is determined that there is a short-circuit precursor in the current single battery, and a short-circuit risk assessment result is obtained, including: Obtain the classification boundary from the short-circuit precursor analysis model and determine the separation threshold between abnormal data and normal data; For the separation threshold, use a sliding window to segment the voltage change rate, temperature sudden rise rate, and internal resistance fluctuation amplitude to obtain a segmented feature sequence; Calculate the statistical values of the features within each segment according to the segmented feature sequence, and obtain the abnormal feature subset; If the statistical value of a feature segment in the abnormal feature subset exceeds the preset threshold, determine the abnormal state through logical judgment to obtain the abnormal state set; Classify the abnormal data using the K-means clustering algorithm through the abnormal state set to obtain the abnormal category set; Judge the status labels corresponding to each abnormal category according to the mapping relationship between the abnormal category set and the operating state to obtain the status classification result; For the status classification result, combine the operating data of the single battery to determine the risk assessment result of the short-circuit precursor.

5. The method according to claim 1, characterized in that, The matching of the pre-established battery pack topology structure diagram with the short-circuit risk assessment result to determine the position of the single battery related to the short-circuit precursor and obtain the target battery identifier to be isolated includes: Obtain the position of the single battery corresponding to the short-circuit risk through the pre-established battery pack topology structure diagram to determine the initial target battery identifier; For the initial target battery identifier, judge the isolation requirement using the preset threshold to obtain the set of batteries to be isolated; From the set of batteries to be isolated, obtain the adjacency relationship of the single battery in the topology to determine the affected battery pack range; According to the affected battery pack range, analyze the propagation path of the short-circuit precursor through the structure diagram to obtain the risk diffusion trend; For the risk diffusion trend, if the position of a battery segment in the propagation path exceeds the preset threshold, determine the high-risk battery subset through logical judgment; From the high-risk battery subset, classify the risk judgment result using the K-means clustering algorithm to obtain the abnormal battery category; According to the abnormal battery category, update the isolation requirement through the topology structure diagram to determine the final set of target battery identifiers.

6. The method according to claim 1, wherein The generation of isolation instructions using the relay control algorithm according to the target battery identifier, and the isolation of the short-circuit battery by disconnecting the corresponding relay connection path to obtain the updated battery pack state includes: Obtain the topology data of the corresponding relay path through the target battery identifier to determine the priority order of the disconnection operation; According to the priority order, generate an isolation instruction sequence using the relay control algorithm to obtain the set of instructions to be executed; For the set of instructions, judge the disconnection condition of the relay path using the preset threshold. If the load of the relay path exceeds the threshold, generate a disconnection signal; From the disconnection signal, obtain the affected connection path range to determine the isolation state of the short-circuit battery; Classify the battery pack state using the K-means clustering algorithm according to the isolation state to obtain the distribution characteristics of the abnormal batteries; Update the topology record of the battery pack state through the distribution characteristics to judge the isolation completion degree of the target battery; Obtain the updated operating parameters of the battery pack from the isolation completion degree, and then obtain the updated battery pack state.

7. The method according to claim 1, wherein Extract the voltage, temperature, and internal resistance data of the remaining single batteries from the updated battery pack state, and adjust the charging current distribution through the equalization management algorithm to obtain the equalized battery operating parameters, including: Obtain the voltage data, temperature data, and internal resistance data of individual cells from the battery pack state, and use an equalization management algorithm to calculate the adjustment value of the charging current to obtain a preliminary current distribution plan; For the preliminary current distribution plan, determine whether the voltage data of the individual cells exceeds the range through a preset threshold. If it exceeds, adjust the charging current to obtain a corrected current distribution result; According to the corrected current distribution result, obtain the change trend of the temperature data of the individual cells, analyze the distribution characteristics of the temperature data using statistical methods, and determine the temperature equalization state; Through the temperature equalization state, obtain the fluctuation range of the internal resistance data, and determine whether the internal resistance data meets the requirements of equalization management to obtain the operating parameters after internal resistance adjustment; Extract the actual distribution value of the charging current from the operating parameters after internal resistance adjustment, and optimize the current distribution using an equalization management algorithm to obtain the balanced battery operating parameters.

8. The method according to claim 1, wherein According to the balanced battery operating parameters, if the voltage, temperature, and internal resistance data of the remaining individual cells are all within the normal range, determine that the system stability is restored, and obtain the final operating state of the battery pack, including: Obtain the voltage difference data of the individual cells from the balanced battery operating parameters, and determine whether the voltage difference meets the range requirements through a preset threshold to obtain a preliminary stability evaluation result; According to the preliminary stability evaluation result, obtain the distribution of the temperature data, analyze the change characteristics of the temperature data using statistical methods, and determine the equalization state of the temperature distribution; Through the equalization state of the temperature distribution, obtain the distribution of the internal resistance data, and determine whether the internal resistance data is within the preset range to obtain the adjustment basis for the internal resistance distribution; For the adjustment basis of the internal resistance distribution, extract the real-time data of the individual cells from the operating parameters to obtain an optimized parameter set; According to the optimized parameter set, obtain the overall change trend of the battery pack state, and determine whether the system stability is restored through logical judgment to obtain a stability confirmation result; Extract the operation conclusion from the stability confirmation result, update the recorded data of the battery pack state, and determine the equalization state of the system operation; Through the recorded data in the equalization state, obtain the long-term operation trend of the individual cells, and determine whether the battery pack state is continuously stable to obtain the final operating state of the battery pack.

9. A computer device, comprising a memory, a processor, and a computer program stored on the memory, characterized in that, The processor executes the computer program to implement the steps of the method according to any one of claims 1-8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the method according to any one of claims 1-8.

Images