Battery cell anomaly identification method, device, equipment, storage medium and program product

Abstract

The application relates to an abnormal cell identification method, device, equipment, storage medium and program product. The method comprises the following steps: collecting parameters of multiple dimensions of a battery cluster in a same string loop at each moment during a charging process; the multiple dimensions at least include voltage, internal resistance and temperature; mapping the parameters of the multiple dimensions at each moment to a three-dimensional feature space constructed according to the voltage, the internal resistance and the temperature; determining discrete data points according to density information of each data point in the three-dimensional feature space; and determining a defective cell according to the discrete data points. The method can more comprehensively capture abnormal features of the cell, thereby improving the accuracy of defective cell identification.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of battery, in particular to a battery cell abnormality identification method and device, computer equipment, computer readable storage medium and computer program product. BACKGROUND

[0002] In the field of energy storage power stations, a battery cluster composed of a large number of battery cells is the core energy storage unit, and its performance stability directly determines the system efficiency and safety. Due to manufacturing differences, environmental fluctuations and other factors, the battery cells will gradually show abnormalities such as capacity attenuation and internal resistance increase. If not identified in time, it will reduce the available capacity of the system, or even cause a thermal runaway accident.

[0003] The current mainstream abnormality identification methods have obvious limitations: the traditional threshold method only judges by the parameter range of voltage, temperature and other parameters, ignores the data distribution rule and change trend, and has insufficient sensitivity; the equivalent circuit model relies on parameters such as internal resistance that are difficult to extract, and has low estimation accuracy, making it difficult to be applied in practice; although the machine learning method has potential, it requires a large amount of labeled data, and the parameter tuning is complex and requires high computing power, making it difficult to adapt to the on-site rapid early warning needs. SUMMARY

[0004] Therefore, it is necessary to provide a battery cell abnormality identification method, device, computer equipment, computer readable storage medium and computer program product capable of improving the accuracy of battery cell abnormality identification.

[0005] In a first aspect, the present application provides a battery cell abnormality identification method, comprising:

[0006] Collecting parameters of multiple dimensions of a battery cluster in the same string loop at each time during the charging process; the multiple dimensions at least include voltage, internal resistance and temperature;

[0007] Mapping the parameters of multiple dimensions at each time to a three-dimensional feature space constructed according to the voltage, the internal resistance and the temperature;

[0008] Determining a discrete data point according to the density information of each data point in the three-dimensional feature space;

[0009] Determining a defective battery cell according to the discrete data point.

[0010] In one of the embodiments, the determining a discrete data point according to the density information of each data point in the three-dimensional feature space comprises:

[0011] For each data point, calculating the local reachable density of the data point and the local reachable density of adjacent data points in the neighborhood of the data point according to a preset neighborhood size;

[0012] calculating a local outlier factor of the data point according to the local reachable density of the data point and the local reachable density of the adjacent data point;

[0013] if the local outlier factor is greater than a predetermined threshold, determining the data point as a discrete data point.

[0014] In one embodiment, before mapping the parameters of multiple dimensions at each time point into a three-dimensional feature space constructed according to the voltage, the internal resistance and the temperature, the method further comprises:

[0015] standardizing the parameters of multiple dimensions at each time point;

[0016] mapping the standardized parameters into the three-dimensional feature space.

[0017] In one embodiment, after collecting the parameters of multiple dimensions of the battery cluster in the same string loop at each time point during the charging process, the method further comprises:

[0018] In any dimension, determining a target time point with the maximum data dispersion degree from the time points based on the parameters of the dimension;

[0019] detecting outliers of the parameters at the target time point to determine the discrete data point.

[0020] In one embodiment, the method further comprises:

[0021] collecting parameters of any dimension of the battery cluster in the same string loop at each time point during two different charging processes;

[0022] determining candidate discrete data points according to the parameters collected in each charging process, respectively;

[0023] taking the intersection of the candidate discrete data points determined in the two times as the discrete data point.

[0024] In one embodiment, the method further comprises:

[0025] According to the temperature of the battery cluster, the temperature change rate of the single battery cell, the temperature change extreme value and the temperature rise per unit of electric quantity are counted.

[0026] According to the temperature change rate, the temperature change extreme value and the temperature rise per unit of electric quantity, a defective battery cell is determined.

[0027] In a second aspect, the application further provides an abnormal battery cell identification device, comprising:

[0028] A data collection module is configured to collect parameters of multiple dimensions of a battery cluster in a same string loop at each time during a charging process, wherein the multiple dimensions at least include voltage, internal resistance and temperature.

[0029] A data mapping module is configured to map the parameters of the multiple dimensions at each time into a three-dimensional feature space constructed according to the voltage, the internal resistance and the temperature.

[0030] A discrete point determination module is configured to determine a discrete data point according to density information of each data point in the three-dimensional feature space.

[0031] An abnormality positioning module is configured to determine a defective battery cell according to the discrete data point.

[0032] In a third aspect, the present application further provides a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the following steps when executing the computer program:

[0033] parameters of multiple dimensions of a battery cluster in a same string loop at each time during a charging process are collected, wherein the multiple dimensions at least include voltage, internal resistance and temperature.

[0034] parameters of multiple dimensions at each time are mapped into a three-dimensional feature space constructed according to the voltage, the internal resistance and the temperature.

[0035] a discrete data point is determined according to density information of each data point in the three-dimensional feature space.

[0036] a defective battery cell is determined according to the discrete data point.

[0037] In a fourth aspect, the present application further provides a computer readable storage medium, which stores a computer program, wherein the computer program is executed by a processor to implement the following steps:

[0038] parameters of multiple dimensions of a battery cluster in a same string loop at each time during a charging process are collected, wherein the multiple dimensions at least include voltage, internal resistance and temperature.

[0039] parameters of multiple dimensions at each time are mapped into a three-dimensional feature space constructed according to the voltage, the internal resistance and the temperature.

[0040] a discrete data point is determined according to density information of each data point in the three-dimensional feature space.

[0041] a defective battery cell is determined according to the discrete data point.

[0042] In a fifth aspect, the present application also provides a computer program product comprising a computer program which, when executed by a processor, implements the following steps:

[0043] Collecting multiple dimensions of parameters of the battery cluster in the same string loop at each time during the charging process; the multiple dimensions at least include voltage, internal resistance and temperature;

[0044] Mapping the multiple dimensions of parameters at each time to a three-dimensional feature space constructed according to the voltage, the internal resistance and the temperature;

[0045] Determining discrete data points according to the density information of each data point in the three-dimensional feature space;

[0046] Determining defective battery cells according to the discrete data points.

[0047] The above battery cell anomaly identification method, device, computer equipment, computer readable storage medium and computer program product collect multiple dimensions of parameters of the battery cluster in the same string loop at each time during the charging process; the multiple dimensions at least include voltage, internal resistance and temperature; map the multiple dimensions of parameters at each time to a three-dimensional feature space constructed according to the voltage, the internal resistance and the temperature; determine discrete data points according to the density information of each data point in the three-dimensional feature space; and determine defective battery cells according to the discrete data points. This method considers three key parameters, voltage, internal resistance and temperature, which reflect the state of the battery cell from different angles. Voltage reflects the electrochemical state of the battery cell, internal resistance reflects the internal loss of the battery cell, and temperature is closely related to the heating and safety of the battery cell. By mapping them to a three-dimensional feature space for analysis, the possible one-sidedness of single parameter detection is avoided, and the abnormal characteristics of the battery cell can be more comprehensively captured, thereby improving the accuracy of defective battery cell identification; further, this method determines discrete data points by analyzing the density of data points in the three-dimensional feature space, without needing to preset a data distribution model, and can adapt to the complex and variable distribution of battery cell parameters in the battery cluster. Whether it is a concentrated distribution of normal battery cell parameters or a dispersed distribution of defective battery cell parameters, abnormal points can be effectively identified through density analysis, enhancing the applicability and robustness of the method. BRIEF DESCRIPTION OF DRAWINGS

[0048] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the related art, the following will briefly introduce the drawings needed to be used in the description of the embodiments of the present application or the related art. Obviously, the drawings in the following description only some embodiments of the present application, and for those skilled in the art, other related drawings can also be obtained without creative labor on the basis of these drawings.

[0049] Figure 1This is a flowchart illustrating a battery cell anomaly identification method in one embodiment;

[0050] Figure 2 This is a schematic diagram of the discrete data point identification process in one embodiment;

[0051] Figure 3 This is a flowchart illustrating a cell anomaly identification method in another embodiment;

[0052] Figure 4 This is a structural block diagram of a battery cell anomaly identification device in one embodiment;

[0053] Figure 5 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0054] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application. It should be noted that the terms "comprising" and "having," and any variations thereof, as used in this application, are intended to cover non-exclusive inclusion.

[0055] In one embodiment, such as Figure 1 As shown, a method for identifying abnormal battery cells is provided. This embodiment illustrates the application of this method to a terminal. It is understood that this method can also be applied to a server, and further to a system including both a terminal and a server, and implemented through interaction between the terminal and the server. In this embodiment, the method includes the following steps:

[0056] Step S110: Collect parameters of multiple dimensions of the battery cluster in the same series circuit at various times during the charging process; the multiple dimensions include at least voltage, internal resistance and temperature.

[0057] In practice, this can be achieved by using voltage sensors, internal resistance testers, and temperature sensors for each individual cell in the series circuit. After the battery cluster begins charging, parameters such as voltage, internal resistance, and temperature for each cell can be collected in real time at a set sampling frequency. During the data collection process, the time information of each collection point must be recorded simultaneously so that the parameters can be correlated with the specific moment later.

[0058] In some embodiments, current parameters of the battery cells can also be acquired using current sensors. The consistency of current data of battery clusters with the same parallel circuit is compared, and data differences are compared by means of mean, range, variance, etc., thereby comparing differences in cluster current circulation, persistent cumulative charge and discharge differences, internal resistance differences, etc.

[0059] Step S120, map the parameters of multiple dimensions at each time to a three-dimensional feature space constructed according to voltage, internal resistance and temperature.

[0060] Exemplarily, a three-dimensional feature space is constructed with voltage as the X-axis, internal resistance as the Y-axis and temperature as the Z-axis. For each single battery cell at each time, the voltage, internal resistance and temperature values are taken as a three-dimensional coordinate point, and the point is mapped to the three-dimensional feature space.

[0061] After the above processing, the parameters of all single battery cells in the battery cluster at each time form a group of data points in the three-dimensional feature space, and the data points at different times constitute a dynamic distribution state in the space.

[0062] Step S130, determine the discrete data points according to the density information of each data point in the three-dimensional feature space.

[0063] In a specific implementation, for each data point in the three-dimensional feature space, the density information of each data point can be calculated by a density calculation method to determine the discrete data points.

[0064] Exemplarily, the k-nearest neighbor density estimation method can be used to calculate the average distance of k nearest neighbor data points around each data point. The smaller the average distance, the greater the density of the region where the data point is located; otherwise, the greater the average distance, the smaller the density. The value of k can be determined according to the data volume and distribution.

[0065] In some embodiments, the change rates of the parameters of each dimension within the same period can be counted, the differences between the battery cells are compared, and the defective battery cells are located.

[0066] Step S140, determine the defective battery cells according to the discrete data points.

[0067] In a specific implementation, after the discrete data points are determined, the battery cells corresponding to the discrete data points are determined as the defective battery cells. In some embodiments, the defective battery cells can be divided into severely defective battery cells and generally defective battery cells according to the discrete degree of the battery cells.

[0068] Processing for severely defective battery cells: if the degree of data deviation of the severely defective battery cell from the average value of normal battery cells is greater than a threshold m, it is determined that the severely defective battery cell must be replaced and its use needs to be stopped as soon as possible; if the deviation degree is in (n, m] (n < m), it is determined that the severely defective battery cell needs to be replaced as soon as possible; if the deviation degree is in [0, n], it is determined that the severely defective battery cell can still be operated.

[0069] For the general defective battery cell: the general defective battery cell is processed in an equalization manner, the degree of deviation of the general defective battery cell from the average value of the normal battery cell voltage is recorded as a, the soc range corresponding to the deviation degree a is obtained by table lookup method, and the equalization time is calculated according to the equalization current size. The battery cell is equalized in the spare time.

[0070] In some embodiments, the number of times or duration that each battery cell is marked as a discrete data point during the entire charging process can be counted. If a certain battery cell is marked as a discrete data point at multiple times, or a discrete data point appears in a critical charging stage (such as a fast charging stage, a near full charge stage), and its discrete degree is relatively significant, it can be determined that the battery cell is a defective battery cell. In some embodiments, historical data of the battery cell can also be analyzed. If the condition of a certain battery cell being determined as a discrete data point in this charging process is similar to the characteristics of previous defective battery cells, it is further confirmed that it is a defective battery cell, thereby improving the accuracy of the determination.

[0071] In the above battery cell anomaly recognition method, the parameters of multiple dimensions of the battery cluster in the same string loop at each time during the charging process are collected; the multiple dimensions at least include voltage, internal resistance and temperature; the multiple dimensions of parameters at each time are mapped into a three-dimensional feature space constructed according to voltage, internal resistance and temperature; the discrete data points are determined according to the density information of each data point in the three-dimensional feature space; and the defective battery cell is determined according to the discrete data points. This method considers three key parameters, voltage, internal resistance and temperature, which reflect the state of the battery cell from different angles. Voltage reflects the electrochemical state of the battery cell, internal resistance reflects the internal loss of the battery cell, and temperature is closely related to the heating and safety of the battery cell. By mapping them into a three-dimensional feature space for analysis, the possible one-sidedness of single parameter detection is avoided, and the abnormal characteristics of the battery cell can be more comprehensively captured, thereby improving the accuracy of the identification of defective battery cells. Further, this method determines the discrete data points by analyzing the density of the data points in the three-dimensional feature space, without needing to preset the distribution model of the data, and can adapt to the complex and variable distribution of the parameters of the battery cells in the battery cluster. Whether it is a concentrated distribution of normal battery cell parameters or a scattered distribution of defective battery cell parameters, abnormal points can be effectively identified through density analysis, enhancing the applicability and robustness of the method.

[0072] In one exemplary embodiment, as shown in Figure 2 Step S130 determines the discrete data points according to the density of each data point in the three-dimensional feature space, including:

[0073] Step S131, for each data point, calculates the local reachable density of the data point and the local reachable density of the adjacent data points in the neighborhood of the data point according to the preset neighborhood size;

[0074] Step S132, calculating the local outlier factor of the data point according to the local reachable density of the data point and the local reachable density of the adjacent data points;

[0075] Step S133, if the local outlier factor is greater than a predetermined threshold, determining the data point as a discrete data point.

[0076] The local reachable density is used to measure the density in the neighborhood of each data point. The greater the local reachable density, the more dense the neighborhood of the data point, and vice versa.

[0077] The local outlier factor reflects the difference in the density of the sample point relative to its neighborhood samples. The greater the value, the more likely the point is an outlier.

[0078] In a specific implementation, for each data point, the adjacent data points in its neighborhood are determined according to a predetermined neighborhood size. The reachable distance of the data points in the neighborhood of the data point to the data point is calculated, and the local reachable density of the data point is obtained by calculating the reciprocal of the average value of the reachable distances of all data points in the neighborhood of the data point to the data point. Similarly, the local reachable densities of all adjacent data points in the neighborhood of the data point are calculated in this way. Then, the average value of the ratio of the local reachable densities of all adjacent data points in the neighborhood of the data point to the local reachable density of the data point itself is calculated as the local outlier factor of the data point. Further, the discrete data points are determined according to the size of the local outlier factor. Specifically, a threshold value can be set, and if the local outlier factor of any data point is greater than the threshold value, the data point is determined as a discrete data point.

[0079] In some embodiments, after determining the local outlier factor, the battery defects can be managed in a hierarchical manner according to the local outlier factor. For example, a battery with a local outlier factor value slightly higher than the normal range may be a "potential defect" (such as slight aging) and needs to be monitored; a battery with a very high local outlier factor value may be a "serious defect" (such as internal short circuit) and needs to be replaced immediately. This hierarchical capability avoids "one-size-fits-all" determination and improves the relevance and economy of the maintenance strategy.

[0080] In this embodiment, whether each data point is an outlier is determined by calculating the local outlier factor of each data point, i.e., by comparing the density difference between each data point and its neighborhood data points, identifying data points with sparse surrounding samples and significant differences from neighbors, and achieving defect detection of the battery. This local density detection method can adapt to various abnormal patterns in the battery system, does not require a pre-set data distribution model, and can reduce missed or false judgments due to complex data distribution.

[0081] In an exemplary embodiment, before mapping the plurality of dimensional parameters at each time point into the three-dimensional feature space constructed according to the voltage, internal resistance and temperature, the step S120 further comprises: normalizing the plurality of dimensional parameters at each time point; and mapping the normalized parameters into the three-dimensional feature space.

[0082] In the embodiment, considering that the dimensions and numerical ranges of the voltage, internal resistance and temperature are different, in order to make the three have equal weights in the three-dimensional feature space, the data needs to be normalized and mapped to the interval [0, 1].

[0083] In an exemplary embodiment, after collecting the plurality of dimensional parameters of the battery cluster in the same string loop at each time point during the charging process, the method further comprises: in any dimension, determining a target time point with the maximum data dispersion degree based on the parameters in the dimension from the time points; and performing outlier detection on the parameters at the target time point to determine a dispersed data point.

[0084] The target time point with the maximum data dispersion degree indicates that the data at the time point during the charging process fluctuates more dramatically. It describes the characteristics of the time point and reflects the overall distribution state of the data at the time point. For example, the voltage of a single cell suddenly changes before and after the time point.

[0085] The dispersed data point refers to an abnormal value in which individual data significantly deviates from the overall distribution at the time point with the maximum data dispersion degree. It describes the characteristics of the data point and is a more extreme individual in the overall dispersed data at the time point.

[0086] In a specific implementation, for any dimension, the data in the dimension during the charging process can be intercepted, and the data dispersion degree at each time point can be calculated using a sliding window, so as to determine the time point with the maximum data dispersion degree. For example, the local standard deviation at each time point can be calculated using a sliding window to represent the data dispersion degree at each time point, and the time point with the maximum local standard deviation is further determined as the target time point. The size of the sliding window can be set based on the amount of collected data. Further, the parameters at the target time point can be subjected to outlier detection using a normal distribution or a Nair detection method to determine the dispersed data point.

[0087] For example, taking the temperature of a cell as an example, the temperature of the cell in the same string loop is collected, the continuous charging and discharging process temperature data is intercepted, the time is recorded as [Tcs, Tce], the differential calculation is performed on the data in the time [Tcs, Tce] to obtain the continuous charging and discharging process temperature differential data. The time point Tcl with the maximum dispersion degree of the charging and discharging process temperature differential data in the time [Tcs, Tce] is detected, and the outliers at the time point Tcl are calculated using a normal distribution or a Nair detection method, which are recorded as {Cellc}. The cells corresponding to the set {Cellc} are the defective cells.

[0088] In this embodiment, the "normal range" of parameters such as voltage, internal resistance and temperature of the battery cell changes with the charging stage (for example, the voltage is low and stable at the beginning of charging, and may fluctuate due to polarization effect in the middle stage). Outliers at a certain time may be normal data at other times. That is, the data anomaly of the battery cell charging process may be the superposition of time dimension anomaly (overall fluctuation at a certain time) and data dimension anomaly (individual extreme value at that time). Therefore, this embodiment adopts the method of first locking the time point with the largest dispersion degree, and then detecting outliers in the local data distribution at that time, thereby ensuring the accuracy of the anomaly judgment.

[0089] In an exemplary embodiment, the method further comprises: collecting parameters of any dimension of the battery cluster in the same string loop at each time in two different charging processes; determining candidate discrete data points according to the parameters collected in each charging process, respectively; and taking the intersection of the candidate discrete data points determined in the two charging processes as the discrete data points.

[0090] In specific matters, for the determination of discrete data points in any dimension, two discrete data point identification operations can be performed, that is, two charging operations are performed, and data in this dimension is collected in each charging process. According to the method of first determining the target time point with the largest data dispersion degree, and then determining the discrete data points at the target time point, the discrete data points obtained in the two charging processes are determined as candidate discrete data points, and the intersection of the candidate discrete data points obtained in the two charging processes is taken as the discrete data points.

[0091] For example, taking the battery cell voltage as an example, the first collection is: collecting the voltage of the battery cell in the same string loop, intercepting the voltage data in the complete charging process, recording the time as [Tcs, Tce], obtaining the time point Tcl with the largest dispersion degree in the time [Tcs, Tce], calculating the outliers at the time point Tcl using the normal distribution or Nair detection method, and recording them as {Cellc}.

[0092] The second collection is: intercepting the voltage data in the complete charging process, recording the time as [Tds, Tde], obtaining the time point Tdl with the largest dispersion degree in the time [Tds, Tde], and calculating the outliers at the time point Tdl using the normal distribution or Nair detection method, and recording them as {Celld}.

[0093] The intersection of {Cellc} and {Celld} is taken as the discrete data points, and the battery cell corresponding to the discrete data points is recorded as the severely defective battery cell.

[0094] In this embodiment, two charging operations are performed to determine two discrete data points, and the intersection of the two determined discrete data points is taken as the final discrete data point, so that the randomness of a single identification result can be avoided, the accuracy of the determined discrete data point is improved, and the positioning accuracy of the defective battery cell is improved.

[0095] In one exemplary embodiment, the method further comprises: according to the temperature of the battery cluster, counting the temperature change rate, temperature change extreme value and temperature rise per unit of electric quantity of the single battery cell; and determining the defective battery cell according to the temperature change rate, temperature change extreme value and temperature rise per unit of electric quantity.

[0096] In a specific implementation, for each single battery cell of the battery cluster, the temperature change rate can be calculated within a preset time window, or the instantaneous change rate can be calculated by using a sliding window. During the charging and discharging process, the highest temperature and the lowest temperature of each single battery cell can be recorded, and the difference is obtained to obtain the temperature change extreme value. The ratio of the temperature change amount to the electric quantity change amount can be counted to determine the temperature rise per unit of electric quantity. Further, the defective battery cell is determined according to the temperature change rate, the temperature change extreme value and the temperature rise per unit of electric quantity.

[0097] In some embodiments, corresponding preset values can be set for the temperature change rate, the temperature change extreme value and the temperature rise per unit of electric quantity. If all three indicators exceed the corresponding preset values, the corresponding battery cell is marked as a defective battery cell.

[0098] In some embodiments, the temperature change rate, the temperature change extreme value and the temperature rise per unit of electric quantity are input into a pre-trained neural network model for processing to obtain the defect probability of the battery cell.

[0099] In some embodiments, the temperature change rate, the temperature change extreme value and the temperature rise per unit of electric quantity are used as features to construct a multi-dimensional vector space, and then a clustering algorithm is used to cluster the data of each battery cell. The battery cell corresponding to the noise point obtained after clustering is taken as a defective battery cell.

[0100] In this embodiment, the temperature-related features (temperature change rate, temperature change extreme value, and temperature rise per unit of electric quantity) of the single battery cells in the battery cluster are analyzed to identify the defective battery cell. The combination of the three temperature indicators can more comprehensively depict the abnormal state of the battery cell, reduce misjudgment, and accurately locate the defective battery cell.

[0101] Reference Figure 3 The flowchart of the battery cell anomaly identification method shown in another embodiment is shown in the flowchart of the battery cell anomaly identification method. In this embodiment, the method comprises the following steps:

[0102] In step S310, the parameters of the battery cluster in the same string loop at each time during the charging process are collected; the multiple dimensions include at least voltage, internal resistance and temperature.

[0103] In step S320, the parameters of the plurality of dimensions at each time are standardized.

[0104] In step S330, the standardized parameters at each time are mapped into a three-dimensional feature space constructed according to the voltage, the internal resistance and the temperature.

[0105] In step S340, for each data point in the three-dimensional feature space, the local reachable density of the data point and the local reachable density of the neighboring data points in the neighborhood of the data point are calculated according to a preset neighborhood size.

[0106] In step S350, the local outlier factor of the data point is calculated according to the local reachable density of the data point and the local reachable density of the neighboring data points.

[0107] In step S360, if the local outlier factor is greater than a predetermined threshold, the data point is determined as a discrete data point.

[0108] In step S370, the defective battery cell is determined according to the discrete data point.

[0109] The method simultaneously considers the three key parameters of voltage, internal resistance and temperature, which reflect the state of the battery cell from different angles. The voltage reflects the electrochemical state of the battery cell, the internal resistance reflects the internal loss of the battery cell, and the temperature is closely related to the heating and safety of the battery cell. By mapping them into a three-dimensional feature space for analysis, the possible one-sidedness of single parameter detection is avoided, and the abnormal features of the battery cell can be more comprehensively captured, thereby improving the accuracy of the identification of the defective battery cell. Further, the method determines the discrete data point by analyzing the density of the data points in the three-dimensional feature space, without needing to preset the distribution model of the data, and can adapt to the complex and variable distribution of the parameters of the battery cells in the battery cluster. Whether the normal battery cell parameters are concentratedly distributed or the defective battery cell parameters are dispersedly distributed, the abnormal points can be effectively identified through density analysis, thereby enhancing the applicability and robustness of the method.

[0110] It should be understood that, although each step in the flowchart involved in each embodiment as described above is shown in sequence according to the arrow, these steps are not necessarily executed in sequence according to the arrow. Unless otherwise explicitly stated herein, the execution of these steps is not strictly limited in sequence, and these steps can be executed in other sequences. Moreover, at least part of the steps in the flowchart involved in each embodiment as described above can include multiple steps or stages, which are not necessarily executed at the same time, but can be executed at different times, and the execution sequence of these steps or stages is not necessarily sequential, but can be executed in rotation or alternation with at least part of other steps or steps or stages in other steps.

[0111] Based on the same inventive concept, the embodiments of the present application also provide a battery cell abnormality identification device for implementing the above-mentioned battery cell abnormality identification method. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme described in the above-mentioned method, and therefore the specific limitations in one or more battery cell abnormality identification device embodiments provided below can refer to the limitations of the battery cell abnormality identification method described above, which will not be repeated here.

[0112] In one exemplary embodiment, as shown in Figure 4 a battery cell abnormality identification device is provided, comprising:

[0113] The data acquisition module 410 is configured to acquire a plurality of dimensional parameters of the battery cluster in the same string loop at each time during the charging process, wherein the plurality of dimensional parameters at least include voltage, internal resistance and temperature.

[0114] The data mapping module 420 is configured to map the plurality of dimensional parameters at each time to a three-dimensional feature space constructed according to the voltage, the internal resistance and the temperature.

[0115] The discrete point determination module 430 is configured to determine a discrete data point according to density information of each data point in the three-dimensional feature space.

[0116] The abnormality positioning module 440 is configured to determine a defective battery cell according to the discrete data point.

[0117] In one embodiment, the discrete point determination module 430 is further configured to, for each data point, calculate a local reachable density of the data point and a local reachable density of a neighboring data point in a neighborhood of the data point according to a preset neighborhood size; calculate a local outlier factor of the data point according to the local reachable density of the data point and the local reachable density of the neighboring data point; and determine the data point as a discrete data point if the local outlier factor is greater than a predetermined threshold.

[0118] In one embodiment, the device further comprises a data processing module configured to perform standardization processing on the plurality of dimensional parameters at each time; and map the standardized parameters to the three-dimensional feature space.

[0119] In one embodiment, the discrete point determination module 430 is further configured to, in any one dimension, determine a target time with the maximum data dispersion degree based on the parameter in the dimension from the times; and perform outlier point detection on the parameter at the target time to determine the discrete data point.

[0120] In one embodiment, the discrete point determination module 430 is further configured to collect parameters of the battery cluster in the same series circuit at any time during two different charging processes; determine candidate discrete data points based on the parameters collected during each charging process; and take the intersection of the two determined candidate discrete data points as the discrete data point.

[0121] In one embodiment, the anomaly location module 440 is further configured to, based on the temperature of the battery cluster, statistically analyze the temperature change rate, extreme temperature change value, and temperature rise per unit charge of individual cells; and to determine defective cells based on the temperature change rate, extreme temperature change value, and temperature rise per unit charge.

[0122] Each module in the aforementioned battery cell anomaly identification device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in the processor of a computer device in hardware form or independent of it, or stored in the memory of a computer device in software form, so that the processor can call and execute the corresponding operations of each module.

[0123] In one exemplary embodiment, a computer device is provided, which may be a terminal, and its internal structure diagram may be as follows: Figure 5 As shown, the computer device includes a processor, memory, input / output interfaces, a communication interface, a display unit, and an input device. The processor, memory, and input / output interfaces are connected via a system bus, and the communication interface, display unit, and input device are also connected to the system bus via the input / output interfaces. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The input / output interfaces are used for exchanging information between the processor and external devices. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, mobile cellular networks, Near Field Communication (NFC), or other technologies. When executed by the processor, the computer program implements a method for identifying abnormal battery cells. The display unit is used to form a visually visible image and can be a display screen, a projection device, or a virtual reality imaging device. The display screen can be an LCD screen or an e-ink screen. The input device of the computer device can be a touch layer covering the display screen, or buttons, trackballs, or touchpads set on the casing of the computer device, or external keyboards, touchpads, or mice, etc.

[0124] Those skilled in the art will understand that Figure 5The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0125] In one embodiment, a computer device is also provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above method embodiments.

[0126] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon that, when executed by a processor, implements the steps in the above method embodiments.

[0127] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above method embodiments.

[0128] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, artificial intelligence (AI) processors, etc., and are not limited to these.

[0129] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0130] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A method for identifying abnormal battery cells, characterized in that, The method includes: During two different charging processes, parameters of the battery cluster in the same series circuit at various times are collected in multiple dimensions; the multiple dimensions include at least voltage, internal resistance and temperature. For each charging process, the following operations are performed: For any dimension, extract the data for that dimension during the charging process, calculate the data dispersion at each moment using a sliding window, and determine the target moment with the largest data dispersion from all the moments, wherein the data dispersion is characterized by local standard deviation; map the parameters of multiple dimensions at the target moment to a three-dimensional feature space constructed based on the voltage, the internal resistance, and the temperature; and determine candidate discrete data points based on the density information of each data point in the three-dimensional feature space. The intersection of the two determined candidate discrete data points is taken as the discrete data point; The number of times each cell in the battery cluster is marked as a discrete data point during the charging process is counted. Cells that are marked as discrete data points at multiple times are identified as defective cells.

2. The method according to claim 1, characterized in that, The step of determining discrete data points based on the density information of each data point in the three-dimensional feature space includes: For each data point, the local reachability density of the data point and the local reachability density of neighboring data points within the neighborhood of the data point are calculated based on a preset neighborhood size. Calculate the local outlier factor of the data point based on the local reachability density of the data point and the local reachability density of the adjacent data points; If the local outlier factor is greater than a predetermined threshold, then the data point is determined to be a discrete data point.

3. The method according to claim 1, characterized in that, Before mapping the parameters of multiple dimensions at the target time to a three-dimensional feature space constructed based on the voltage, the internal resistance, and the temperature, the method further includes: The parameters of multiple dimensions at the target time are standardized. The standardized parameters are mapped to the three-dimensional feature space.

4. The method according to claim 1, characterized in that, The method further includes: Based on the temperature of the battery cluster, the temperature change rate, extreme temperature change value, and temperature rise per unit charge of individual cells are statistically analyzed. Defective cells are identified based on the rate of temperature change, the extreme value of temperature change, and the temperature rise per unit charge.

5. The method according to claim 1, characterized in that, After identifying the defective battery cell, the process also includes: Based on the degree of dispersion of the defective cells, the defective cells are classified into severely defective cells or generally defective cells.

6. A battery cell anomaly identification device, characterized in that, The device includes: The data acquisition module is used to collect parameters of multiple dimensions of the battery cluster in the same series circuit at various times during two different charging processes; the multiple dimensions include at least voltage, internal resistance and temperature. The data mapping module is used to extract data for any dimension during each charging process, calculate the data dispersion at each moment using a sliding window, determine the target moment with the largest data dispersion from all the moments, and characterize the data dispersion by local standard deviation; and map the parameters of multiple dimensions at the target moment to a three-dimensional feature space constructed based on the voltage, the internal resistance, and the temperature. The discrete point determination module is used to determine candidate discrete data points based on the density information of each data point in the three-dimensional feature space, and to take the intersection of the two determined candidate discrete data points as the discrete data point. An anomaly location module is used to count the number of times each cell in the battery cluster is marked as a discrete data point during the charging process, and to identify defective cells that are marked as discrete data points at multiple times.

7. The apparatus according to claim 6, characterized in that, The discrete point determination module is further configured to, for each data point, calculate the local reachability density of the data point and the local reachability density of neighboring data points within the neighborhood of the data point according to a preset neighborhood size; calculate the local outlier factor of the data point based on the local reachability density of the data point and the local reachability density of the neighboring data points; and determine the data point as a discrete data point if the local outlier factor is greater than a predetermined threshold.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 5.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 5.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 5.

Images