Battery pack internal short circuit fault detection method and system based on multi-feature fusion

Abstract

The application discloses a battery pack internal short circuit fault detection method and system based on multi-feature fusion. The method comprises the following steps: collecting voltage data, current data and probe temperature data of each single battery in the battery pack at real time; calculating an average maximum voltage difference F1 according to the voltage data; calculating a maximum decentralized voltage deviation F2 after the voltage data is processed by decentralization; calculating a weighted absolute current-temperature rise ratio F3 according to the current data and the probe temperature data; integrating the average maximum voltage difference F1, the maximum decentralized voltage deviation F2 and the weighted absolute current-temperature rise ratio F3 into feature points, and integrating and mapping the feature points into data points; clustering each data point by using a DBSCAN clustering algorithm to obtain each abnormal data point, and obtaining the battery single of the battery pack internal short circuit according to each abnormal data point. The application collects voltage, current and temperature data in the battery pack, extracts three features to represent the battery pack state, and can obviously improve the detection accuracy and detection timeliness.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of lithium ion battery early warning, and particularly relates to a battery pack internal short circuit fault detection method based on multi-feature fusion. BACKGROUND

[0002] Lithium ion batteries are currently widely used in electric vehicles and energy storage systems, but internal short circuits of battery packs occur from time to time. If the internal short circuit of the battery pack cannot be monitored in time and effectively, the battery management system (BMS) does not make timely response strategies, and extreme conditions such as thermal runaway and even fire of the battery pack are extremely likely to occur. Therefore, it is necessary to detect the internal short circuit of the battery pack.

[0003] The internal short circuit detection algorithms of lithium ion batteries mainly include gas detection, temperature rise detection, and data-driven methods. In recent years, domestic and foreign scholars have designed corresponding monitoring systems and mathematical models for the internal short circuit problem of lithium ion batteries. There are usually no gas sensors in the lithium ion battery pack in actual deployment, so the method using gas sensors is often for laboratory and other scenarios. For the temperature rise detection method, some scholars have designed a three-level early warning system for a certain type of lithium ion battery, and 50℃, 70℃ and 80℃ are used as the three-level early warning temperatures. This method of monitoring only using a certain feature inevitably encounters problems such as sensor data distortion and even failure in the actual operation of lithium ion batteries, and it is difficult to ensure the early warning speed by using only temperature data to predict internal short circuits. Some scholars have also predicted battery internal short circuits based on charging data, but in reality, many battery packs have internal short circuits not only during charging, but also during discharging and frequent charging and discharging. Although many internal short circuit detection algorithms for lithium ion batteries have been proposed, they are usually only applicable to a single working condition or laboratory conditions. In the actual operation of energy storage systems and electric vehicle battery packs, the battery pack is usually in frequent charging and discharging, so it is difficult to cover most of the use conditions by only predicting internal short circuits during charging. Moreover, the working environment of lithium ion batteries is not as good as that of laboratories, and the number of lithium ion batteries in the battery pack is often large, so data loss and distortion are inevitable. In addition, as the battery pack ages and its health state decreases, the electrical performance parameters of the battery and its sensors decrease, resulting in low accuracy and long detection time for battery pack internal short circuit detection, which leads to short early warning time. SUMMARY

[0004] To solve the above technical problems, the application provides a battery pack internal short circuit fault detection method and system based on multi-feature fusion.

[0005] In a first aspect, the application provides a battery pack internal short circuit fault detection method based on multi-feature fusion, which comprises:

[0006] S1: real-time interval sampling the voltage data, current data and probe temperature data of each monomer in the battery pack;

[0007] S2: calculating the average maximum voltage difference F1 according to the voltage data of each monomer in the battery pack;

[0008] S3: calculating the maximum decentralization voltage deviation F2 according to the voltage data of each monomer in the battery pack after decentralization processing;

[0009] S4: calculating the weighted absolute current-temperature rise ratio F3 according to the current data and probe temperature data of each monomer in the battery pack;

[0010] S5: integrating the average maximum voltage difference F1, the maximum decentralization voltage deviation F2 and the weighted absolute current-temperature rise ratio F3 into feature points, and mapping the feature points into data points;

[0011] S6: clustering each data point obtained by mapping by using a DBSCAN clustering algorithm to obtain each abnormal data point, and obtaining the short-circuit battery monomer in the battery pack according to each abnormal data point.

[0012] In a second aspect, the present application provides a battery pack internal short circuit fault detection system based on multi-feature fusion, which comprises:

[0013] A data acquisition module is configured to collect voltage data, current data and probe temperature data of each monomer in the battery pack in real time at intervals in response to a fault detection request;

[0014] A data processing module is configured to process the voltage data, current data and probe temperature data of each monomer to obtain an average maximum voltage difference F1, a maximum decentralization voltage deviation F2 and a weighted absolute current-temperature rise ratio F3, and to normalize the same to obtain corresponding features after normalization;

[0015] A data integration module is configured to map the corresponding features after normalization into data points;

[0016] A clustering module is configured to cluster each data point obtained by mapping by using a DBSCAN clustering algorithm to obtain each abnormal data point;

[0017] An output module is configured to output a battery pack internal short circuit fault detection result according to each abnormal data point.

[0018] In a third aspect, the present application provides an electronic device comprising:

[0019] One or more processors;

[0020] a memory for storing one or more instructions, wherein the one or more instructions, when executed by the one or more processors, cause the one or more processors to implement the method according to the first aspect of the present application.

[0021] In a fourth aspect, the present application provides a computer-readable storage medium having stored thereon executable instructions that, when executed by a processor, cause the processor to perform the method according to the first aspect of the present application.

[0022] The present application has the following beneficial effects: the present application uses the sliding window method, uses the voltage, current and temperature data in the BMS system in actual work, extracts three features of average maximum voltage difference, maximum decentralization deviation and weighted absolute current-temperature rise ratio to represent the state of the battery pack, can well score whether there is an internal short circuit in the battery pack, and then uses the DBSCAN unsupervised clustering method to accurately find out the abnormal time point and locate the internal short circuit battery monomer, which overcomes the false alarm problem caused by partial sensor error when only using a single feature method, improves the fault detection speed and accuracy, provides valuable decision-making time for the BMS system, and greatly improves the safety performance of the BMS system and the thermal management system. BRIEF DESCRIPTION OF DRAWINGS

[0023] Figure 1 A step flowchart of the battery pack internal short circuit fault detection method in the embodiment of the present application;

[0024] Figure 2 A flowchart of the battery pack internal short circuit fault detection method in the embodiment of the present application;

[0025] Figure 3 A structure diagram of the battery pack internal short circuit fault detection system in the embodiment of the present application. DETAILED DESCRIPTION

[0026] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work fall within the scope of protection of the present application.

[0027] The present application has long-term research on existing BMS data, and it is found that lithium ion battery pack will have occasional frame loss, distortion and other situations during operation. Moreover, with the aging of the battery pack and BMS, the electrical performance parameters of various sensors in the battery will also decrease, resulting in higher system error. If it cannot be effectively identified, the false positive rate and false negative rate of the detection result will be high, and it will take longer time to get the detection result (i.e. the warning time left for BMS becomes shorter).

[0028] The embodiment of the present application proposes a battery pack internal short circuit fault detection method based on multi-feature fusion, as shown in Figure 2 、 Figure 3 The method comprises:

[0029] S1: real-time interval sampling of voltage data, current data and probe temperature data of each battery monomer in the battery pack.

[0030] Specifically, the battery pack is composed of an array of N battery monomers (referred to as monomers). During operation, the battery pack generates various types of operation data in real time, and the collected data is continuously fed back to the battery management system (BMS) through sensors. For example, the operating parameters of the battery pack in an electric vehicle include voltage, current, battery temperature, battery capacity, battery energy, state of health (SOH) and the like. The real-time voltage, current and battery temperature of each battery monomer in the battery pack are obtained through voltage sensors, current sensors and temperature sensors, and are fed back to the battery management system (BMS).

[0031] In response to a fault detection request, the voltage data, current data and probe temperature data of each battery monomer in the battery pack are sampled at intervals from the BMS using a sliding window method, i.e. M frames of voltage data, current data and probe temperature data are selected, and at least S data window segments are selected.

[0032] S2: calculating the average maximum voltage difference according to the voltage data of each monomer in the battery pack.

[0033] Specifically, in the sampling segment, the maximum voltage difference between each monomer is calculated according to the voltage data of each monomer in the battery pack, and the average value of the maximum voltage difference between each battery monomer is calculated to obtain the average maximum voltage difference.

[0034] The average maximum voltage difference, i.e. the average value of the maximum voltage difference between the monomers in the sampling segment, is calculated as follows:

[0035] S201: in the sampling segment, the voltage matrix of each battery monomer in the time window (i.e. the voltage matrix of the battery pack) is obtained according to the voltage data of each monomer in the battery pack, which is specifically represented as:

[0036]

[0037] Further,

[0038] wherein V represents the voltage matrix within the time window, V i, represents the single cell voltage row vector of all batteries at the i-th moment, V ,j represents the all voltage column vector of the j-th single cell within the time window, u i,j represents the voltage of the j-th single cell at the i-th moment, represents the matrix transpose, M represents the length of the sliding window, and N represents the number of battery single cells.

[0039] S202: Find the voltage value of each row from the voltage matrix V within the time window, and calculate the maximum single cell voltage difference between each battery single cell in each row:

[0040] V i_max_diff = max (u i,j - u i,k ), i∈[1, M]; j, k∈[1, N]

[0041] wherein V i_max_diff represents the maximum single cell voltage difference between the battery single cells at the i-th moment, max(.) represents the maximum value, u i,j represents the voltage of the j-th single cell at the i-th moment, i represents the i-th moment, j represents the j-th single cell, and k represents the k-th single cell.

[0042] S203: Calculate the average value of the maximum voltage difference according to V i_max_diff of each row in the voltage matrix within the time window:

[0043]

[0044] wherein V max_diff_mean represents the maximum voltage difference within the time window. V max_diff_mean is represented by F1, i.e., F1 = V max_diff_mean .

[0045] In a battery pack, due to manufacturing process, material, use condition and other factors, there will be certain performance differences between each battery monomer. These differences will cause inconsistent monomer voltages during the operation of the battery pack, especially during charging and discharging. If these differences are not managed and adjusted, the voltage of each monomer battery will gradually differentiate, and the service life will be greatly reduced. In view of the battery equalization management, the voltage values of each battery monomer are decentralized processed. Compared with the existing voltage value decentralization processing, the embodiment of the present application is not simply subtracting the matrix mean value, but subtracting the row mean value first, and then subtracting the column mean value. The row mean value is subtracted to filter the voltage changes at different time points caused by different load conditions, and the column mean value is subtracted to filter the physical differences between different monomers, so as to better capture the voltage drop of the same monomer at different times caused by external factors such as non-load, and then more accurate voltage characteristic values can be extracted.

[0046] S3: According to the voltage data of each monomer in the battery pack, the maximum decentralized voltage deviation is calculated after the decentralized processing.

[0047] Specifically: in the sampling segment, according to the voltage data of each monomer in the battery pack, the voltage values of each battery monomer are decentralized processed, and the maximum decentralized voltage deviation is calculated.

[0048] The process of calculating the maximum decentralized voltage deviation is specifically:

[0049] S301: Obtain the voltage matrix V of each battery monomer in the time window, calculate the relative difference value of the voltage matrix V and the row mean value, and obtain the row mean difference matrix V diff .

[0050] Specifically, in the sampling segment, according to the voltage data of each monomer in the battery pack, the relative difference value of the voltage matrix window and the row mean value is calculated, and the row mean difference matrix V diff between the voltage matrix and the row mean value can be obtained, which is specifically:

[0051]

[0052] Wherein, V represents the voltage matrix in the time window, V row_mean represents the row mean value of the voltage matrix V, that is, all elements equal to the mean value of the row where they are located; V diff represents the row mean difference matrix of the voltage matrix V minus the row mean value V rowmean , V i, represents the monomer voltage row vector of all batteries at the i-th moment, V i,_mean represents the row vector of the average value of the monomer voltage of all batteries at the i-th moment, and the length is N.

[0053] S302: Calculate the voltage matrix V and obtain the column mean V based on the voltage data of each cell in the battery pack. col_mean Calculate the row mean difference matrix V diff With the mean of column V of the voltage matrix V col_mean The relative difference is used to obtain the decentralized matrix of the individual unit voltage.

[0054] The specific calculation process for S302 is as follows:

[0055] V dece_volt =V diff -V col_mean ,

[0056]

[0057] and then,

[0058] Among them, V dece_volt V represents the decentralized matrix of individual unit voltages. diff V represents the matrix obtained by subtracting the row mean from V. col_mean This indicates that the voltage matrix V takes the column mean, meaning that all elements in it are equal to the mean of their respective columns. ,j_diff V represents diff The j-th column, V ,j_mean V represents diff The j-th single entity V in this window for all time periods diff The average value is represented by a column vector of length M.

[0059] The decentralized matrix of individual cell voltage reflects the voltage change value within a window after excluding the influence of the cell's own physical properties.

[0060] S303: Based on the individual voltage decentralization matrix, the maximum decentralization voltage deviation is calculated by subtracting the average decentralization voltage of the individual voltage from the average decentralization voltage of the individual voltage in the last 10 frames within the time window.

[0061] Specifically, the mean (V) of the decentralized voltage of each individual unit within the first 10 frames of the window is used. dece_volt (j)[1:10]), minus the mean (V) of the decentralized voltage of each individual in the last 10 frames within the window. dece_volt (j)[M-9:M]), obtain the decentralized voltage deviation of the single cell within this window, and find its maximum decentralized voltage deviation V. max_dece_diff :

[0062] V dece_diff (j)=mean(V dece_volt (j)[1:10])-mean(V dece_volt (j)[M-9:M])

[0063] V max_dece_diff = max(V dece_diff )

[0064] where j represents the jth battery cell in the battery pack, [x:y] represents selecting the xth to yth frame data, and V max_dece_diff is represented by F2, i.e., F2 = V max_dece_diff .

[0065] The temperature-related detection in a lithium ion battery is currently usually directly set to a threshold value for judgment, and if the threshold value is exceeded, it is considered abnormal. However, in a cold environment, in the early stage of a short circuit in the battery, although the temperature in the battery is also rapidly rising, it cannot reach the fault threshold point in time. At the same time, if the battery itself is undergoing large-current discharge and charging, the temperature rise of the battery will also be faster. Therefore, the present application considers this weighted absolute current-temperature rise ratio feature to determine whether the battery pack has an abnormal temperature rise.

[0066] S4: According to the current data and probe temperature data of each cell in the battery pack, the weighted absolute current-temperature rise ratio is calculated.

[0067] Specifically, in the sampling segment, according to the voltage data and probe temperature data of each cell in the battery pack, the current value of each frame and the probe temperature difference between every two frames in each battery cell are obtained, and the weighted absolute current-temperature rise ratio is calculated.

[0068] The weighted absolute current-temperature rise ratio can reflect the proportion of the current battery temperature rise and work. Under normal circumstances, the battery will only have a temperature rise when it is working and charging. If there is an abnormal temperature rise, this feature can capture it well.

[0069] The weighted absolute current-temperature rise ratio calculation method is as follows:

[0070] S401: According to the current data of each cell in the battery pack, the absolute value average of the current between two frames in the window is calculated to obtain the absolute current average, which is specifically as follows:

[0071] First, according to the current data of each cell in the battery pack, the battery pack total current vector in the sliding window is calculated, which is represented as:

[0072]

[0073] where C represents the battery pack total current matrix in the time segment, C i represents the battery pack total current at i time.

[0074] Then, the absolute current average between two frames in the time window is calculated, which is specifically as follows:

[0075]

[0076] where C represents the current vector within the time segment, length M, C avg represents the absolute current average value between two frames of current within the window, length M-1, C[x:y] represents the vector composed of the xth frame to the yth frame in C.

[0077] S402: According to the probe temperature data of each monomer in the battery pack, the probe temperature difference between every two frames within the window is calculated, which is specifically:

[0078] First, according to the probe temperature data of each monomer in the battery pack, the temperature matrix within the time window is obtained, which is specifically:

[0079]

[0080] Further,

[0081] where T represents the temperature matrix within the time window, T i, represents the temperature row vector of all temperature probes at the ith moment, T ,j represents the temperature column vector of the jth temperature probe within the current time window, t i,j represents the temperature value of the jth temperature probe at the ith moment, (.) T represents the matrix transpose, M represents the length of the sliding window, and N represents the number of temperature probes.

[0082] Then, the probe temperature difference between every two frames within the window is calculated, which is specifically:

[0083] T ,j_diff = T ,j [2:M]-T ,j [1:M-1]

[0084]

[0085] where T ,j_diff represents the temperature difference value vector of the jth temperature probe within the time window, T ,j [x:y] represents the vector composed of the xth to yth value of T ,j , T diff represents the matrix composed of each temperature probe after calculating the temperature difference value vector, size (M-1)*S, T i,_diff represents the ith row element of T diff , which is the row vector composed of the temperature rise value of all temperature probes at the ith~(i+1) time. j represents the temperature probe in the jth battery pack, and S represents the number of temperature probes in the battery pack.

[0086] S403: Calculate the probe maximum temperature difference between each two frames according to the probe temperature difference between each two frames in the window.

[0087] The embodiment of the present application uses the maximum temperature rise sequence to represent the probe maximum temperature difference, which is specifically:

[0088]

[0089] Wherein, T max_diff represents the maximum temperature rise sequence, max(.) represents the maximum value, max(T i,_diff ) represents the maximum value of the temperature rise value of all temperature probes at the i~(i+1) time.

[0090] S404: Calculate the weighted absolute current-temperature rise ratio sum WS according to the absolute current average value and the probe maximum temperature difference.

[0091] Specifically, the element product of the absolute current average value and the probe maximum temperature difference is calculated, and then divided by the time interval, and then weighted summation is performed to obtain the weighted absolute current-temperature rise ratio sum.

[0092] In the embodiment of the present application, the weight is set as the number of calculated ratios, which is specifically:

[0093]

[0094] Wherein, WS represents the weighted absolute current-temperature rise ratio sum, i is the iteration variable and the set weight, the closer to the time point of the sliding window, the higher the weight value, T max_diff (i) represents the i-th element in T max_diff , that is, max(T i,_diff ), |. | is the absolute value, C avg (i) represents the i-th element in C avg , in order to avoid calculation errors or abnormal weighted absolute current-temperature rise ratio sum caused by too small current value, when |C avg (i)|<1, |C avg (i)|=1, Δt(i) represents the time interval from the i-th frame to the i+1-th frame in the sliding window.

[0095] S405: Calculate the weighted absolute current-temperature rise ratio according to the weighted absolute current-temperature rise ratio sum WS and the sliding segment length, and the calculation formula is:

[0096] F3=WAER,

[0097]

[0098] Further,

[0099] wherein, WAER represents a weighted absolute current-temperature rise ratio, WAER is represented by F3, i represents a weight of each iteration, and M represents a sliding window length.

[0100] S5: integrating the average maximum voltage difference F1, the maximum decentralized voltage deviation F2, and the weighted absolute current-temperature rise ratio F3 into a feature point (F1, F2, F3), and mapping the integration into a data point.

[0101] Specifically, in a three-dimensional Euclidean coordinate, the average maximum voltage difference F1, the maximum decentralized voltage deviation F2, and the weighted absolute current-temperature rise ratio F3 are integrated into a feature point (F1, F2, F3), and the feature point (F1, F2, F3) is integrated and mapped into a data point.

[0102] After multiple repeated sampling and processing through steps S1-S5, a feature-integrated data point set is obtained. To study the law of the data point set, the DBSCAN algorithm is used in the embodiment of the present application.

[0103] Density-based clustering method with noise (Density-Based Spatial Clustering of Applications with Noise, DBSCAN) is a density-based spatial clustering algorithm. The algorithm divides the region with sufficient density into clusters and discovers clusters of arbitrary shape in a spatial database with noise. It defines a cluster as the maximum set of density-connected points.

[0104] DBSCAN is a scale-sensitive clustering algorithm, so normalization processing needs to be performed on the three features (F1, F2, and F3).

[0105] Further, the average maximum voltage difference F1, the maximum decentralized voltage deviation F2, and the weighted absolute current-temperature rise ratio F3 are normalized, and the normalized features are integrated into a feature point.

[0106] The calculation formula of the normalization processing is:

[0107]

[0108] wherein, F scaled represents the normalized feature, F represents the original value of the feature, F min represents the minimum value of the feature, and F max represents the maximum value of the feature.

[0109] After step S5, the data points mapped by the feature integration include the features of voltage, current and temperature rise in the battery pack. The diagnosis of the internal short circuit in the battery in the present application no longer depends on a single feature, which reduces the misdiagnosis rate, and fully utilizes the correlation of the three features, so that the internal short circuit in the battery pack can be detected as early as possible.

[0110] S6: clustering each data point obtained by mapping by using a DBSCAN clustering algorithm to obtain each abnormal data point, and obtaining the battery monomer with an internal short circuit in the battery pack according to each abnormal data point.

[0111] Select appropriate DBSCAN hyperparameters ε, and define the radius MinPts of the neighborhood: the minimum number of neighborhood points required for a point to be considered as a core point.

[0112] DBSCAN clustering is performed on each data point, and the specific clustering steps are as follows:

[0113] S601: mark all data points as unprocessed;

[0114] S602: for each unprocessed point P, obtain all points in the ε neighborhood of point P, denoted as set N, and then check the core point, if the number of points in set N is greater than or equal to MinPts, mark point P as a core point, otherwise temporarily mark the data point as a boundary point;

[0115] S603: for each point Q in set N, if point Q is an unprocessed point, obtain the ε neighborhood N(Q) of point Q, if the number of points in neighborhood N(Q) is greater than or equal to MinPts, mark point Q as a core point, and add all points in neighborhood N(Q) to the current cluster; if point Q does not belong to any cluster, temporarily mark it as a boundary point;

[0116] S604: repeat S602 and S603 until all data points are processed, and determine the data points that are not added to any cluster as boundary points, which are the abnormal data points.

[0117] After finding the abnormal data points, enter the data corresponding to the segment, find the battery module corresponding to the temperature probe with the largest temperature rise, and find the monomer with the largest voltage drop, that is, the battery monomer with the internal short circuit.

[0118] Based on the same inventive concept, the present application embodiment proposes a battery pack internal short circuit fault detection system based on multi-feature fusion. In this system, it has the same or similar technical features as the above-mentioned battery pack internal short circuit fault detection method, and the following will not be repeated.

[0119] Referring to Figure 3 The battery pack internal short circuit fault detection system based on multi-feature fusion includes:

[0120] The data acquisition module is configured to collect voltage data, current data and probe temperature data of each single battery in the battery pack in real time at a time interval in response to a fault detection request.

[0121] The data processing module is configured to process the voltage data, current data and probe temperature data of each single battery to obtain average maximum voltage difference F1, maximum decentralization voltage deviation F2 and weighted absolute current-temperature rise ratio F3, and normalize the F1, F2 and F3 to obtain corresponding features after normalization.

[0122] The data integration module is configured to integrate and map the corresponding features after normalization into data points.

[0123] The clustering module is configured to cluster each data point obtained by mapping by using a DBSCAN clustering algorithm to obtain abnormal data points.

[0124] The output module is configured to output a short-circuit fault detection result in the battery pack according to the abnormal data points.

[0125] An electronic device is provided in an embodiment of the present application, and the electronic device comprises:

[0126] One or more processors;

[0127] A memory is configured to store one or more instructions, and when the one or more instructions are executed by the one or more processors, the one or more processors implement the method provided in the first aspect of the present application.

[0128] A computer readable storage medium is provided in an embodiment of the present application, and the computer readable storage medium stores executable instructions, and when the executable instructions are executed by a processor, the processor implements the method provided in the first aspect of the present application.

[0129] The present application has the following beneficial effects: the present application uses a sliding window method, extracts voltage, current and temperature data that can be extracted in an actual BMS system in work, extracts three features (F1, F2 and F3) and integrates the three features into a feature point to represent the state of the battery pack, and the battery pack can be scored well to determine whether there is an internal short circuit; subsequently, a DBSCAN unsupervised clustering method is used to accurately find abnormal segments, which overcomes the false positives caused by partial sensor errors when only a single feature is used, and also takes into account the early warning speed.

[0130] The present application can provide valuable decision-making time for a BMS system and greatly improve the safety performance of the BMS system and a thermal management system.

[0131] Those skilled in the art can understand that all or part of the steps in the above-mentioned embodiments can be completed by programs instructing relevant hardware, and the programs can be stored in a computer readable storage medium, which can include ROM, RAM, magnetic disk or optical disk, etc.

[0132] Although the embodiments of the present application have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made therein without departing from the principles and spirit of the application, and the scope of the present application is defined by the appended claims and their equivalents.

Claims

1. A battery pack internal short circuit fault detection method based on multi-feature fusion, characterized in that, The method comprises: real-time interval sampling of voltage data, current data and probe temperature data of each single cell in the battery pack; The average maximum voltage difference is calculated from the voltage data of the individual cells in the battery pack ; According to the voltage data of each single body in the battery pack, the maximum decentralization voltage deviation is calculated after decentralization processing ; Based on the current data and the probe temperature data of each single body in the battery pack, a weighted absolute current-temperature rise ratio is calculated ; normalizing the average maximum voltage difference , the maximum decentralized voltage deviation , and the weighted absolute current-temperature rise ratio , integrating the normalized features into feature points, and mapping the integration of the feature points into data points; clustering each data point obtained by mapping using a DBSCAN clustering algorithm to obtain each abnormal data point, and obtaining a short-circuit battery cell in the battery pack according to each abnormal data point.

2. The battery pack internal short circuit fault detection method based on multi-feature fusion according to claim 1, characterized in that, The calculation method of the average maximum voltage difference is specifically: in the sampling segment, the maximum voltage difference between each single cell is solved according to the voltage data of each single cell in the battery pack, the average value of the maximum voltage difference between each battery cell is calculated, and the average maximum voltage difference is obtained.

3. The battery short circuit fault detection method based on multi-feature fusion according to claim 1, characterized in that, The calculation process of the average maximum voltage difference is specifically: Within the sampled segment, a voltage matrix of each battery cell within the time window is acquired ; calculating the voltage matrix maximum cell voltage difference between each cell in each row ; The average maximum voltage difference is calculated from the average of the maximum voltage differences of each row in the voltage matrix. , calculating the average of the maximum voltage differences, resulting in an average maximum voltage difference.

4. The battery short circuit fault detection method based on multi-feature fusion according to claim 1, characterized in that, The calculation process of the maximum decentralized voltage deviation is: Within the sampled segment, a voltage matrix of each battery cell within the time window is acquired ; computing the voltage matrix the relative difference from its row mean, resulting in a row mean difference matrix ; computing the row mean difference matrix with the voltage matrix column mean the relative difference, resulting in a monomer voltage decentering matrix ; According to the monomer voltage decentralization matrix, using the average value of the decentralized voltage of the previous 10 frames in the time window Subtract the average value of the decentralized voltage of the last 10 frames of the monomer voltage , the maximum decentralized voltage deviation is calculated , wherein represents the first battery monomer in the battery pack, represents the length of the sliding window.

5. The battery short circuit fault detection method based on multi-feature fusion according to claim 1, characterized in that, The calculation process of the weighted absolute current-temperature rise ratio is: According to the current data of each single cell in the battery pack, the absolute current average value between two frames in the time window is calculated; According to the probe temperature data of each single cell in the battery pack, the probe temperature difference between every two frames in the time window is calculated, and the probe maximum temperature difference between every two frames is calculated according to the probe temperature difference; According to the absolute current average value and the probe maximum temperature difference, a weighted absolute current-temperature rise ratio and ; According to the weighted absolute current-temperature rise ratio and and the slide fragment length, the weighted absolute current-temperature rise ratio is calculated.

6. The battery internal short circuit fault detection method based on multi-feature fusion according to claim 1, characterized in that, The clustering process of each data point obtained by mapping using the DBSCAN clustering algorithm is specifically: Set the DBSCAN hyperparameter ε to define the radius of the neighborhood MinPts; S601: mark all data points as unprocessed; S602: for each unprocessed point P, obtain all points in the ε neighborhood of point P, denoted as set N, then check the core point, if the number of points in set N is greater than or equal to MinPts, mark point P as a core point, otherwise temporarily mark the data point as a boundary point; S603: for each point Q in set N, if point Q is an unprocessed point, obtain the ε neighborhood N(Q) of point Q, if the number of points in neighborhood N(Q) is greater than or equal to MinPts, mark point Q as a core point, and add all points in neighborhood N(Q) to the current cluster; if point Q does not belong to any cluster, temporarily mark it as a boundary point; S604: repeat S602 and S603 until all data points are processed, and the data points that are not added to any cluster are determined as boundary points, which are abnormal data points.

7. A battery internal short circuit fault detection system based on multi-feature fusion, the system is based on the battery internal short circuit fault detection method based on multi-feature fusion as claimed in claim 1, characterized in that, The system comprises: a data acquisition module configured to collect voltage data, current data and probe temperature data of each single cell in the battery pack in real time at an interval in response to a fault detection request; a data processing module configured to process voltage data, current data, and probe temperature data of each single cell to obtain average maximum voltage difference , maximum decentralization voltage deviation , and weighted absolute current-temperature rise ratio , and normalize the data to obtain corresponding features after normalization, respectively. a data integration module configured to integrate and map the corresponding features processed by normalization into data points; a clustering module configured to cluster each data point obtained by mapping using a DBSCAN clustering algorithm to obtain each abnormal data point; an output module configured to output a short-circuit fault detection result in the battery pack according to each abnormal data point. 8.An electronic device comprising: one or more processors; a memory configured to store one or more instructions, wherein the one or more instructions, when executed by the one or more processors, cause the one or more processors to implement the method of any one of claims 1 to 6. 9.A computer-readable storage medium having stored thereon executable instructions that, when executed by a processor, cause the processor to implement the method of any one of claims 1 to 6.

Images