Battery pack abnormal cell detection method and system based on local outlier factor algorithm

Abstract

The application discloses a battery pack abnormal monomer detection method and system based on a local outlier factor algorithm. The method comprises the following steps: sampling and collecting voltage data and current data of each battery monomer in a battery pack by using a sliding window method to obtain a monomer voltage matrix; first voltage features F1 and second voltage features F2 are obtained according to the monomer voltage matrix U; a voltage change consistency feature F3 of each monomer in the sliding window is calculated according to the monomer voltage matrix U and the current data of the battery monomer; the first voltage features F1, the second voltage features F2 and the voltage change consistency feature F3 are integrated into feature points, and the feature points are integrated and mapped into data points; a local outlier factor LOF algorithm is used to perform abnormal detection on the data point set to obtain abnormal data points, and the abnormal data points are abnormal battery monomers in the battery pack. The application can improve the accuracy of abnormal monomers in the battery pack and can judge the type of abnormal conditions.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of electric vehicle battery management, and particularly relates to a battery pack abnormal single cell detection method and system based on a local outlier factor algorithm. BACKGROUND

[0002] With the development of new energy technology, power battery packs are increasingly widely used, such as electric vehicles, energy storage devices, electric tools, and the like. A power battery pack includes multiple battery cells in series or in parallel. Generally, power battery cell inconsistency refers to that in the same battery pack, each battery cell has significant differences in parameters such as voltage, capacity, internal resistance, temperature characteristics, and the like, which are beyond the normal range, resulting in a decline in the overall performance of the battery pack and constituting a safety hazard. The intuitive manifestations of power battery cell inconsistency are as follows: one is that the power battery has inconsistent cell performance parameters such as internal resistance and capacity, and the other is that the power battery has inconsistent working states such as working voltage and state of charge. At present, the methods for identifying the inconsistency of cells in an electric vehicle battery pack mainly include machine learning, data clustering analysis, and outlier detection

[0003] The basic idea of power battery cell inconsistency identification through machine learning is to extract features, train a machine learning model through relevant features, and then use the trained model to identify abnormal cells. Some scholars use wavelet analysis to extract a diagnostic feature vector and involve a fault diagnosis BP neural network for diagnosing power battery cell inconsistency. The disadvantage of machine learning algorithms is that the calculation amount is large, and the battery management system (BMS) of an electric vehicle is not suitable for algorithms with a large amount of data calculation, so it is difficult to deploy in the power battery BMS system. In addition, different hyperparameters need to be set for different data sets, and different hyperparameter settings often bring completely different results, so it is difficult to apply to different vehicle models.

[0004] The basic idea of the method based on data clustering analysis is to extract feature points, then perform clustering analysis on each feature point, and thus find out abnormal points and identify abnormal cells. However, in some general clustering algorithms, the number of clusters needs to be specified in advance. This method needs to know the types of power battery cell inconsistency before clustering. Some scholars have proposed a power battery voltage inconsistency identification algorithm based on dynamic k-value Kmeans++ clustering by optimizing the number of clustering clusters and the selection of initial centers on the basis of the K-means clustering algorithm. Although this method can identify potential abnormal cells to a certain extent, the number of clusters still needs to be specified in advance.

[0005] Currently, the algorithm based on outlier detection is usually through mathematical statistics method, such as calculating Z-score using 3-sigma criterion for judgment, or making box plot, identifying abnormal value through quartile method, etc., for identifying data points which are significantly different from most data. The dimension requirement of such method to data cannot be too high, and the traditional outlier detection method often cannot effectively process high-dimensional data, at the same time, the noise and abnormal fluctuation in data may affect the calculation result, resulting in that a large number of normal data are misjudged as abnormal. SUMMARY

[0006] To solve the problems in the prior art, the application provides a battery pack abnormal single cell detection method and system based on a local outlier factor algorithm.

[0007] In a first aspect, the application provides a battery pack abnormal single cell detection method based on a local outlier factor algorithm, which comprises:

[0008] S1: using a sliding window method, sampling and collecting voltage data and current data of each battery cell in the battery pack to obtain a single cell voltage matrix;

[0009] S2: calculating an average Z-score of each battery cell in the sliding window according to the single cell voltage matrix U to obtain a first voltage feature F1;

[0010] S3: calculating a second voltage feature F2 of each single cell in the sliding window according to the single cell voltage matrix U;

[0011] S4: calculating a voltage change consistency feature F3 of each single cell in the sliding window according to the single cell voltage matrix U and the current data of the battery cell;

[0012] S5: integrating the first voltage feature F1, the second voltage feature F2 and the voltage change consistency feature F3 into a feature point, and integrating and mapping the feature point into a data point;

[0013] S6: using a local outlier factor LOF algorithm to perform abnormal detection on the data point set to obtain an abnormal data point, which is an abnormal battery cell in the battery pack.

[0014] In a second aspect, the application provides a battery pack abnormal single cell detection system based on a local outlier factor algorithm, which comprises:

[0015] A data acquisition module, configured to, in response to a fault detection request, collect voltage data and current data of each single cell in a battery pack at a real-time interval;

[0016] A feature extraction module, configured to perform data processing on the voltage data and current data of the single cells, and extract features to obtain a first voltage feature F1, a second voltage feature F2 and a voltage change consistency feature F3.

[0017] characteristics F1, the second voltage characteristic F2, and the voltage change consistency characteristic F3 into a feature point, and map the integration of the feature point into a data point to obtain a data point set of each battery monomer of the battery pack;

[0018] The detection module is internally provided with a local outlier factor (LOF) algorithm program, and is used for performing anomaly detection on the data point set to determine the type of the abnormal condition of the abnormal point.

[0019] The detection module is internally provided with a local outlier factor (LOF) algorithm program, and is used for performing anomaly detection on the data point set to determine the type of the abnormal condition of the abnormal point.

[0020] In a third aspect, the present application provides an electronic device, comprising:

[0021] one or more processors;

[0022] a memory for storing one or more instructions, wherein when the one or more instructions are executed by the one or more processors, the one or more processors implement the method proposed in the first aspect of the present application.

[0023] In a fourth aspect, the present application provides a computer-readable storage medium having executable instructions stored thereon, wherein the executable instructions are executed by a processor to make the processor implement the method proposed in the first aspect of the present application.

[0024] The present application has the following beneficial effects: by combining the sliding window method, three different characteristics (F1, F2 and F3) are calculated, which can well score each battery in the last time window, and after normalization of the characteristics, different scales of the characteristics are normalized to the same dimension, so as to facilitate subsequent local outlier factor (LOF) calculation, and by selecting appropriate k value and LOF threshold value, the abnormal point can also be accurately found for further analysis, which not only overcomes the dependence of the traditional clustering method on the number of specified clusters, but also does not require a large amount of calculation related to machine learning algorithms. BRIEF DESCRIPTION OF DRAWINGS

[0025] Figure 1 The step flowchart of the embodiment of the present application;

[0026] Figure 2 The overall flowchart of the embodiment of the present application;

[0027] Figure 3 The structure diagram of the battery pack abnormal monomer detection system in the embodiment of the present application;

[0028] Figure 4 The LOF score diagram of each monomer in the battery pack during simulation experiment of the present application;

[0029] Figure 5 Figure 2 is a diagram of the difference between each single cell voltage and average voltage in the battery pack during simulation experiments of the present application. DETAILED DESCRIPTION

[0030] The terms "first", "second", "third", and the like in the description and claims of the present application and the above figure are used for distinguishing between similar objects and not necessarily for describing a specific sequential or chronological order. It is to be understood that the terms so used are interchangeable under appropriate circumstances and are merely employed in the description of embodiments of the present application for descriptive purposes.

[0031] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without creative work fall within the protection scope of the present application.

[0032] The embodiments of the present application provide a battery pack abnormal single cell detection method and system based on a local outlier factor algorithm, and the method comprises the following steps:

[0033] S1: The voltage data and current data of each battery cell in the battery pack are sampled and collected by using a sliding window method, to obtain a single cell voltage matrix.

[0034] For example, there are N battery cells (which can be referred to as single cells) in a target battery pack (which can also be referred to as a battery group). The size of the sliding window is set, and M frames of voltage data of each battery cell in the battery pack are collected.

[0035] Through research and analysis of the existing battery management system (BMS) data set of an electric vehicle, it is found that the BMS data of the electric vehicle during operation has certain distortion and sampling random error. If such short error fragments cannot be effectively identified, the normal battery cell will be easily marked as an abnormal battery cell. The sliding window filtering method is used in the embodiments of the present application, which can effectively filter abnormal data and retain the original characteristics of the data.

[0036] The single cell voltage matrix U is specifically as follows:

[0037]

[0038] Wherein, u 1,1 represents the voltage data of the first battery cell in the battery pack in the first frame in the sliding window, u 1,N represents the voltage data of the Nth battery cell in the battery pack in the first frame in the sliding window, u M,1 represents the voltage data of the first battery cell in the battery pack in the Mth frame in the sliding window, uM,N represents the voltage data of the Nth battery monomer in the battery pack in the sliding window at the Mth frame.

[0039] Through the research and analysis of the abnormal battery monomer data in the battery pack of a large number of electric vehicles, it is found that the voltage data of the abnormal battery monomer is different from the voltage average, the voltage median, and the sensitivity to the current from the ordinary battery monomer, and therefore, the method is designed to extract the Z-score of the battery monomer in the sliding window, the consistency coefficient, and the product PVMD of the monomer voltage standard deviation and the median absolute deviation, and the three features are taken as three-dimensional space points to calculate the local outlier factor of all monomers. Finally, the outlier factor is obtained as a further inconsistency discrimination standard.

[0040] S2: According to the monomer voltage matrix U, the average Z-score of each battery monomer in the sliding window is calculated to obtain a first voltage feature F1.

[0041] Specifically, the Z-score of each monomer voltage data in the sliding window is calculated through the monomer voltage matrix U. First, the Z-score of each frame is calculated, and then the average Z-score of each monomer in the segment is calculated. The calculation formula is specifically:

[0042]

[0043] Wherein, i represents the i-th time, j represents the j-th monomer; Z i,j represents the Z-score of the j-th monomer at the i-th frame, u i,j represents the voltage data of the j-th monomer at the i-th time, represents the average voltage data of the battery pack at the i-th time, std(u i,1 ,…,u i,N ) represents, represents the average Z-score of each battery monomer in the time window, and M represents the number of frames of the time window. The first voltage feature F1 is represented by That is:

[0044] It should be noted that the Z-score (Z-score) is also called standard score (standard score), which is used to describe the relative position of a numerical value and the average. Specifically, it is calculated by dividing the difference between the numerical value and the average by the standard deviation. In the field of data anomaly value detection, Z-score can be used to detect abnormal values in a data set. Generally, if the absolute value of the Z-score is greater than a certain threshold (such as 2 or 3), the data point is considered to be an abnormal value.

[0045] The embodiment of the application calculates the average Z-score of the monomer voltage in the entire segment in the sliding window, which plays a filtering error role and avoids the problem of false positives caused by distortion of a certain frame.

[0046] S3: According to the monomer voltage matrix U, the second voltage feature F2 of each monomer in the sliding window is calculated.

[0047] Specifically, according to the monomer voltage matrix U, the monomer voltage standard deviation and the median absolute deviation of each monomer in the sliding window are calculated respectively, and the product PVMD of the monomer voltage standard deviation and the median absolute deviation is obtained to obtain the second voltage feature F2.

[0048] The specific calculation formula of the product PVMD of the monomer voltage standard deviation and the median absolute deviation is:

[0049]

[0050] PVMD j =std(u ,j )×median(|u ,j -median(U,axis=1)|);

[0051] F2=PVMD j ;

[0052] In the formula, PVMD j represents the product of the monomer voltage standard deviation and the median absolute deviation of the jth monomer, and F2 represents the second voltage feature PVMD j ; u ,j represents the column vector of all voltage data of the jth monomer in the i-th time window, u i, represents the monomer voltage of all battery monomers at the i-th frame, u i,j represents the voltage data of the jth monomer at the i-th moment, std(.) represents the calculation of the standard deviation, median(.) represents the median value of., and if there is an axis parameter, it means that the median value is taken in the corresponding axis direction, such as median(U,axis=1) represents taking the row mean of the matrix U, axis=1 means calculating on the row to obtain an M×1 column vector, wherein each median(|u i,j -median(u i )|) represents the median absolute deviation of the jth monomer in the sliding window.

[0053] The monomer voltage standard deviation indicates the change range of the monomer in the time window, and if the change range is large, the monomer voltage standard deviation will be large

[0054] The monomer with large voltage deviation in the battery pack will affect the voltage mean value, so the median value represented by the battery monomer often better reflects the normal state of the battery pack than the average value. The median absolute deviation indicates the deviation of the monomer from the median value of the voltage in the battery pack in the time window.

[0055] The product of the standard deviation of the cell voltage and the median absolute deviation, which can reflect both the change range of the cell in the time window and the deviation from the median value, and adding this feature can help to enhance the sensitivity of the subsequent local outlier factor algorithm to abnormal cells.

[0056] S4: According to the cell voltage matrix U and the current data of the battery cells, the voltage change consistency feature F3 of each cell in the sliding window is calculated.

[0057] In a preferred embodiment, the calculation process of the voltage change consistency feature F3 is as follows:

[0058] S401: Remove the first frame of the cell voltage matrix U from the cell voltage matrix U to obtain a first voltage matrix U1.

[0059] Specifically, the first voltage matrix U1 is specifically as follows:

[0060]

[0061] S402: Remove the last frame of the cell voltage matrix U from the cell voltage matrix U to obtain a second voltage matrix U2.

[0062] Specifically, the second voltage matrix U2 is specifically as follows:

[0063]

[0064] S403: Subtract the first voltage matrix U1 from the second voltage matrix U2 to obtain the voltage difference matrix U of each cell between every two frames in the sliding window. diff .

[0065] Specifically, the calculation formula of the voltage difference matrix U diff is as follows:

[0066] U diff = U2-U1;

[0067]

[0068] In the formula, u diff i,j represents the voltage difference of the jth cell from i to i+1 time.

[0069] S404: According to the voltage difference matrix U diff and the current data of the battery cells, the Pearson correlation coefficient r of each cell in the sliding window is calculated.

[0070] In some preferred embodiments, according to the voltage difference matrix U diff , the voltage difference matrix Udiff The average value of each row is obtained as the average voltage difference U of each frame of data diff_mean The total current of the entire battery pack is calculated according to the current data of each battery cell in the battery pack; and the Pearson correlation coefficient of each battery cell in the sliding window is calculated according to the voltage difference matrix U diff The average value of each column is obtained as the average voltage difference U of each frame of data diff_mean The total current of the entire battery pack, and the Pearson correlation coefficient of each battery cell in the sliding window is calculated.

[0071] Specifically, the calculation formula of the Pearson correlation coefficient is:

[0072]

[0073] In the formula, I j+1 represents the total current of the battery pack at the i+1 moment, u diff i,j represents the voltage difference of the jth battery cell from the i moment to the i+1 moment.

[0074] It should be noted that the Pearson correlation coefficient (Pearson correlation coefficient) is also called Pearson product-moment correlation coefficient (Pearson product-moment correlation coefficient), which is a statistical quantity for measuring the degree of linear correlation between two variables.

[0075] S405: According to the Pearson correlation coefficient r of each battery cell in the sliding window, the relative deviation RD of the Pearson correlation coefficient of each battery cell is calculated to obtain a consistency coefficient feature F3.

[0076] Specifically, according to the Pearson correlation coefficient r of each battery cell in the sliding window, the average value r of the Pearson correlation coefficient of all battery cells in the entire battery pack is calculated mean , and the relative deviation RD of the Pearson correlation coefficient of each battery cell is calculated.

[0077] The relative deviation of the Pearson correlation coefficient of each battery cell is calculated, and the calculation formula is:

[0078]

[0079] F3=RD j

[0080] In the formula, RD j represents the relative deviation of the Pearson correlation coefficient of the jth battery cell, RD j is represented by F3; r j represents the Pearson correlation coefficient of the jth battery cell in the sliding window.

[0081] In the embodiment of the present application, the Pearson correlation coefficient is used to study the linear relationship between the battery cell voltage and the battery pack current. After the processing of S401 to S405, the voltage variation consistency feature is extracted, which reflects the consistency of the variation of each cell in the battery pack and the battery pack current. By calculating the relative deviation and comparing the differences between different cells, this feature can more sensitively capture the inconsistency of the battery cells caused by the large change in the internal resistance of the battery cells.

[0082] In a preferred embodiment, the first voltage feature F1, the second voltage feature F2, and the consistency coefficient feature F3 are normalized respectively, and the normalized calculation formula is:

[0083]

[0084] In the formula, F scaled represents the normalized feature, F represents the input feature, which is specifically F1, F2, or F3, F min represents the minimum value in the input feature, F max represents the maximum value in the input feature.

[0085] S5: The first voltage feature F1, the second voltage feature F2, and the voltage variation consistency feature F3 are integrated into a feature point, and the feature point integration is mapped into a data point.

[0086] Specifically, in a three-dimensional Euclidean coordinate, the first voltage feature F1, the second voltage feature F2, and the consistency coefficient feature 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.

[0087] Using the same method, the data point corresponding to each cell in the target battery pack is obtained, and a data point set is formed. To study the rule of the data point set, the local outlier factor (LOF) algorithm is used to detect abnormal data points, so that the abnormal cells in the battery pack are detected.

[0088] S6: The local outlier factor LOF algorithm is used to perform anomaly detection on the data point set, and an abnormal data point is obtained, which is an abnormal battery cell in the battery pack.

[0089] In some preferred embodiments, the local outlier factor LOF algorithm is used to perform anomaly detection on the data point set (i.e., to detect abnormal data points), and the specific process includes:

[0090] S601: Determine the number of neighbors k of each data point p, i.e., determine the value of k.

[0091] S602: Calculate the reachable distance reach_dist(p, o) of each data point p and the points in its k-neighborhood.

[0092] Specifically, for each data point p, calculate the points in its k-neighborhood, k-neighborhood N k (p) refers to the k nearest data points to the data point p.

[0093] The formula for calculating the reachable distance reach_dist(p, o) is:

[0094] reach_dist(p, o) = max(dist(p, o), dist(p, N k (p))

[0095] Where N k (p) represents the k-neighborhood of data point p, dist(p, o) represents the Euclidean distance from data point p to any neighbor point o, and dist(p, N k (p)) represents the Euclidean distance from data point p to its k-neighborhood N k (p).

[0096] S603: Calculate the local reachable density lrd(p) of each data point, and the formula is:

[0097]

[0098] Where |N k (p)| represents the size of the k-neighborhood of any data point p, i.e., the value of k.

[0099] Similarly, calculate the local reachable density lrd(o) of any point o in the k-neighborhood of data point p.

[0100] S604: Obtain the local outlier factor LOF score of the data point by comparing the local reachable density lrd(p) of the data point and the local reachable density lrd(o) of the data points in its k-neighborhood.

[0101] Specifically, the formula for calculating the local outlier factor LOF score is:

[0102]

[0103] Where LOF(p) represents the local outlier factor LOF score of any data point p, lrd(p) represents the local reachable density of any data point p, lrd(o) represents the local reachable density of any neighbor data point o in the k-neighborhood of any data point p, and o ∈ N k (p) represents any neighbor data point o in the k-neighborhood of any data point p.

[0104] S605: Set a threshold value, compare the local outlier factor LOF score of the data point p with the threshold value, and obtain an abnormal data point, which is an abnormal battery monomer.

[0105] Specifically, set the LOF score threshold, compare the local outlier factor LOF score of any data point p with the LOF score threshold, and further judge the data points that exceed the LOF score threshold; then, in combination with the monomer voltage matrix U, perform inconsistent state classification and grading, which can be divided into voltage inconsistency fault and internal resistance inconsistency fault. Similarly, in order to avoid sampling anomalies, sliding window processing needs to be performed on the data.

[0106] In a preferred embodiment, when the local outlier factor LOF score of the data point p exceeds the threshold value, in combination with the monomer voltage matrix U, the data point p is detected as an abnormal data point, and its abnormal condition is divided into voltage abnormality (i.e., voltage inconsistency abnormality) and internal resistance abnormality (i.e., internal resistance inconsistency abnormality).

[0107] Specifically,

[0108] For voltage inconsistency abnormality (or fault): when the local outlier factor LOF score of the data point p exceeds the threshold value, calculate the average value of the difference between the monomer voltage and the average voltage in the sliding window, and the calculation formula is:

[0109]

[0110] V diff (i) value reflects the degree of deviation of the voltage of the i-th monomer from the average voltage within the time window.

[0111] Specifically: when V diff (i)≥30mV, it should be considered as monomer voltage inconsistency;

[0112] When V diff (i)≥50mV, it should be considered as moderate monomer voltage inconsistency;

[0113] When V diff (i)≥80mV, it should be considered as severe monomer voltage inconsistency.

[0114] For internal resistance inconsistency abnormality (or fault): when the local outlier factor LOF score of the data point p exceeds the threshold value, for monomers with larger internal resistance, when the current changes, the voltage change caused by the internal resistance part is also larger, which is reflected in the monomer voltage. Its monomer voltage change will be more severe. Therefore, first, the time window segment with continuous time and current change needs to be intercepted, and the standard deviation of the monomer voltage in the time window is calculated, and the calculation formula is:

[0115]

[0116] Subsequently, the average standard deviation of each battery cell in the battery pack within its time window is calculated, and the calculation formula is:

[0117]

[0118] When std(u j )>3×std(u mean ), it should be considered as inconsistent internal resistance of the cell;

[0119] When std(u j )>4×std(u mean ), it should be considered as moderate internal resistance inconsistency of the cell;

[0120] When std(u j )>5×std(u mean ), it should be considered as severe internal resistance inconsistency of the cell.

[0121] Based on the same inventive concept, the embodiment of the application proposes a battery pack abnormal cell detection system based on a local outlier factor algorithm. In this system, it has the same or similar technical features as the above-mentioned battery pack abnormal cell detection method, and the following is not repeated.

[0122] Referring to Figure 3 , the battery pack abnormal cell detection system based on the local outlier factor algorithm comprises:

[0123] A data acquisition module is configured to collect voltage data and current data of each cell in the battery pack at real time intervals in response to a fault detection request;

[0124] A feature extraction module is configured to process the voltage data and current data of each cell, extract features, and obtain first voltage features F1, second voltage features F2, and voltage change consistency features F3;

[0125] A feature integration module is configured to integrate the first voltage features F1, the second voltage features F2, and the voltage change consistency features F3 into feature points, and map the feature points into data points to obtain a data point set of each battery cell in the battery pack;

[0126] A detection module is configured to set a local outlier factor LOF algorithm program therein, and is configured to perform abnormal detection on the data point set to determine the abnormal condition type of the abnormal point;

[0127] The detection module is configured to output and display the abnormal battery cell and the abnormal condition type thereof.

[0128] The embodiment of the application proposes an electronic device, comprising:

[0129] One or more processors;

[0130] 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.

[0131] The embodiment of the present application provides a computer readable storage medium, which stores executable instructions, and the executable instructions are executed by a processor to enable the processor to implement the method according to the first aspect of the present application.

[0132] Simulation verification:

[0133] Based on a computer and a python programming language, the present application uses battery data actually existing battery inconsistency to perform simulation verification, constructs a related algorithm designed by the present application, reads battery pack data, and finally obtains experimental results as follows:

[0134] Figure 4 The LOF score graph of each battery monomer in the battery pack in the simulation experiment of the present application. Figure 4 In the simulation experiment of the present application, the LOF score value calculated by the algorithm is used to monitor the operation data of the electric vehicle. In this detection, the k value in the LOF local outlier factor is set to 5, and the threshold value is set to 10. From Figure 4 It can be seen from the simulation experiment of the present application that the local outlier factor of the 24th battery monomer reaches 16.86 at this moment.

[0135] From the simulation experiment of the present application, the automatic extraction of three characteristics and the calculation of the LOF local outlier factor are used to automatically locate the battery monomer with inconsistency. Figure 4

[0136] The difference graph of the voltage of each battery monomer and the average voltage in the battery pack in the simulation experiment of the present application. Figure 5 In the simulation experiment of the present application, when the battery monomer LOF is abnormal, the difference between the voltage of the battery monomer in the sliding window and the average voltage is used to monitor the operation data of the electric vehicle. From Figure 5 It can be seen from the simulation experiment of the present application that the 24th battery monomer has a larger voltage drop in the operation process, has inconsistency in the voltage drop of the battery monomer, and has a larger voltage fluctuation, and also has inconsistency in the increase of the internal resistance of the battery monomer. Figure 5 From the simulation experiment of the present application, it can be seen that the 24th battery monomer indeed has inconsistency, and unlike the method of directly detecting the voltage of the battery monomer, the method is sensitive not only to the voltage drop of the battery, but also to the increase of the internal resistance of the battery, proving the sensitivity of the algorithm to the imbalance of the battery monomer.

[0137] Figure 5

[0138] ​​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.

[0139] 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 abnormal cell detection method based on a local outlier factor algorithm, characterized in that, The method comprises the following steps: The voltage data and current data of each battery cell in the battery pack are sampled and collected by using a sliding window method to obtain a cell voltage matrix; According to the monomer voltage matrix U, the average Z score of each battery monomer in the sliding window is calculated to obtain a first voltage feature ; According to the cell voltage matrix U, the cell voltage standard deviation and the median absolute deviation of each cell in the sliding window are calculated respectively, and the product of the cell voltage standard deviation and the median absolute deviation is obtained , to obtain the second voltage feature ; Subtracting the first frame of the cell voltage matrix U from the cell voltage matrix U, a first voltage matrix U is obtained Subtracting the last frame of the cell voltage matrix U from the cell voltage matrix U, a second voltage matrix U is obtained ; subtracting the second voltage matrix U from the first voltage matrix U , a voltage difference matrix U is obtained , which is a voltage difference matrix of each cell between every two frames within the sliding window ; calculating the average value of each row in the voltage difference matrix U , an average voltage difference of each frame of data is obtained The total current of the entire battery pack is calculated according to the current data of each battery cell in the battery pack; According to the voltage difference matrix The average voltage difference per column, per frame data And the total current of the entire battery pack, the Pearson correlation coefficient of each monomer in the sliding window is calculated ; Based on the Pearson correlation coefficient of each monomer within the sliding window The relative deviation RD of the Pearson correlation coefficient for each battery cell is calculated to obtain the consistency coefficient characteristics. ; the first voltage characteristic , a second voltage characteristic , a voltage change uniformity characteristic into a feature point, and mapping the feature point integration into a data point; The data point set is detected for abnormality by using a local outlier factor (LOF) algorithm to obtain an abnormal data point, which is an abnormal battery cell in the battery pack.

2. The battery pack abnormal cell detection method based on the local outlier factor algorithm according to claim 1, characterized in that, The method further comprises the following steps: normalizing the first voltage characteristic , a second voltage characteristic , a consistency coefficient characteristic respectively.

3. The battery pack abnormal cell detection method based on a local outlier factor algorithm according to claim 1, characterized in that, The data point set is detected for abnormality by using a local outlier factor (LOF) algorithm, and the specific process comprises the following steps: The number k of neighbors of the data point p is determined, that is, the value of k is determined; Computing the reachable distance of a data point p from data points o in its k-neighborhood ; The local reachable density of the data point p and the data points o in the k-neighborhood of the data point p are calculated respectively; by comparing the local reachable density of data point p with the local reachable density of data point o within the k-neighborhood of p a local outlier factor LOF score of data point p is derived A threshold value is set, the local outlier factor (LOF) score of the data point p is compared with the threshold value, and an abnormal data point is obtained, which is an abnormal battery cell.

4. The battery pack abnormal cell detection method based on a local outlier factor algorithm according to claim 3, characterized in that, When the local outlier factor (LOF) score of the data point p exceeds the threshold value, the data point p is detected as an abnormal data point in combination with the cell voltage matrix U, and the abnormal condition of the data point p is classified into voltage abnormality and internal resistance abnormality.

5. A battery pack abnormal cell detection system based on a local outlier factor algorithm, the system being used to implement the battery pack abnormal cell detection method based on a local outlier factor algorithm according to claim 1, characterized in that, The system comprises: A data collection module for collecting voltage data and current data of each cell in the battery pack at real time intervals in response to a fault detection request; The feature extraction module is configured to perform data processing on the voltage data and the current data of the single units, extract features, and obtain first voltage features , second voltage features , and voltage change consistency features . a feature integration module configured to integrate the first voltage feature , the second voltage feature , and the voltage variation consistency feature into a feature point, and map the feature point integration into a data point to obtain a data point set of each battery monomer of the battery pack; A detection module in which a local outlier factor (LOF) algorithm program is arranged, for detecting abnormality of a data point set and judging the abnormal condition type of the abnormal point; The detection module is used for outputting and displaying the abnormal battery cell and the abnormal condition type thereof. 6.An electronic device comprising: one or more processors; a memory for storing one or more instructions, wherein when the one or more instructions are executed by the one or more processors, the one or more processors implement the battery pack abnormal cell detection method based on the local outlier factor algorithm according to any one of claims 1 to 4. 7.A computer readable storage medium having stored thereon executable instructions that, when executed by a processor, cause the processor to implement the battery pack abnormal cell detection method based on the local outlier factor algorithm according to any one of claims 1 to 4.

Images