Battery sampling anomaly identification method and device, storage medium and electronic equipment

Abstract

The application discloses a battery sampling abnormality identification method and device, a storage medium and electronic equipment. The method comprises the following steps: using a battery management system to sample a plurality of target battery monomers included in a target battery pack to obtain measured voltage data corresponding to the plurality of target battery monomers respectively; using a target neural network model to identify the measured voltage data corresponding to the plurality of target battery monomers respectively to determine an abnormality identification result of the target battery pack; in the case that the abnormality identification result indicates that there is a sampling voltage abnormality, determining voltage time sequence trends corresponding to the plurality of target battery monomers respectively based on the measured voltage data corresponding to the plurality of target battery monomers respectively; and positioning a sampling abnormality event to occur in the battery management system and / or the target battery pack based on the voltage time sequence trends corresponding to the plurality of target battery monomers respectively. The application solves the technical problem of insufficient sampling abnormality identification and positioning capability in the related art.

Description

Technical Field

[0001] This invention relates to the field of battery technology, and more specifically, to a method, apparatus, storage medium, and electronic device for identifying abnormal battery sampling. Background Technology

[0002] The power battery is an indispensable core component of new energy vehicles (electric vehicles), directly affecting the overall vehicle performance and safety; its importance is self-evident. During operation, the battery pack's state parameters (voltage, current, temperature, etc.) are collected, monitored, analyzed, and alerted in real time to ensure the battery pack's health and normal operation. However, due to manufacturing processes, poor contact in the data acquisition lines, and battery pack malfunctions, abnormal voltage acquisition of adjacent cells can occur. This can impact the management strategies of the Battery Management System (BMS), the alerting strategies of cloud-based monitoring algorithms, and the battery health status analysis algorithms, leading to analysis errors, false alarms, and fault detection failures. Therefore, fault location and troubleshooting are necessary. However, current technologies lack the capability to pinpoint the location of the anomaly after it has been identified. Furthermore, the difficulty of disassembling battery packs after assembly further complicates troubleshooting.

[0003] There is currently no effective solution to the above problems. Summary of the Invention

[0004] This invention provides a method, apparatus, storage medium, and electronic device for identifying battery sampling anomalies, in order to at least solve the technical problem of insufficient sampling anomaly identification and location capabilities in related technologies.

[0005] According to one aspect of the present invention, a battery sampling anomaly identification method is provided, comprising: using a battery management system to sample multiple target battery cells included in a target battery pack to obtain measured voltage data corresponding to each of the multiple target battery cells; using a target neural network model to identify the measured voltage data corresponding to each of the multiple target battery cells to determine anomaly identification results of the target battery pack, wherein the target neural network model is trained using a training set, the training set being generated based on voltage feature data corresponding to multiple reference battery packs, the multiple reference battery packs being in different states; when the anomaly identification results indicate the presence of a sampling voltage anomaly, determining the voltage time-series trend corresponding to each of the multiple target battery cells based on the measured voltage data corresponding to each of the multiple target battery cells; and locating the sampling anomaly event occurring in the battery management system and / or the target battery pack based on the voltage time-series trend corresponding to each of the multiple target battery cells.

[0006] Optionally, the method further includes: for any battery pack among the plurality of reference battery packs, obtaining voltage feature data of the arbitrary battery pack; expanding the voltage feature data of the arbitrary battery pack to obtain expanded samples corresponding to the arbitrary battery pack; using the plurality of reference battery packs as the arbitrary battery pack respectively to obtain expanded samples corresponding to the plurality of reference battery packs respectively; and labeling the expanded samples corresponding to the plurality of reference battery packs respectively to obtain the training set.

[0007] Optionally, the step of expanding the voltage characteristic data of the arbitrary battery pack to obtain the expanded sample corresponding to the arbitrary battery pack includes: based on a predetermined sliding step size and a predetermined window timing length, using a sliding window to repeatedly truncate the voltage characteristic data of the arbitrary battery pack to obtain multiple sets of window data with overlap; if the timing length of any set of data in the multiple sets of window data is less than the predetermined window timing length, splicing the arbitrary set of data so that the timing length of the arbitrary set of data is equal to the predetermined window timing length.

[0008] Optionally, the step of locating the sampling anomaly event occurring in the battery management system and / or the target battery pack based on the voltage time-series trends corresponding to the plurality of target battery cells includes: locating the sampling anomaly event occurring in the target battery pack when there are two adjacent battery cells among the plurality of target battery cells, the voltage time-series trends corresponding to the two adjacent battery cells are symmetrically distributed about the mean change trend, and the voltage time-series trend of one of the two adjacent battery cells is higher than the mean change trend, while the voltage time-series trend of the other battery cell is lower than the mean change trend; wherein the mean change trend is the average voltage trend of the other cells among the plurality of target battery cells excluding the two adjacent cells.

[0009] Optionally, the step of locating the sampling anomaly event occurring in the battery management system and / or the target battery pack based on the voltage timing trends corresponding to the plurality of target battery cells includes: in the plurality of target battery cells, there are three adjacent battery cells, namely a first cell, a second cell, and a third cell, wherein the voltage timing trends corresponding to the first cell and the third cell are the same and higher than the average change trend, and the voltage timing trend of the second cell is lower than the average change trend, in which case the sampling anomaly event is located in the battery management system, wherein the average change trend is the average voltage trend of the other cells in the plurality of target battery cells excluding the three adjacent cells.

[0010] Optionally, the step of locating the sampling anomaly event occurring in the battery management system and / or the target battery pack based on the voltage time-series trends corresponding to the plurality of target battery cells includes: if there are two adjacent battery cells among the plurality of target battery cells, and the voltage time-series trends corresponding to the two adjacent battery cells are simultaneously higher or lower than the average change trend, then locating the sampling anomaly event occurring in the target battery pack and the battery management system, wherein the average change trend is the average voltage trend of the other cells among the plurality of target battery cells excluding the two adjacent cells.

[0011] Optionally, after determining that a sampling anomaly event occurred in the battery management system and / or the target battery pack based on the voltage time-series trends corresponding to the plurality of target battery cells, the method further includes: determining a plurality of maintenance battery packs matching the type of the target battery pack, and sampling anomaly location data corresponding to the plurality of maintenance data packets, wherein the plurality of maintenance battery packs are battery packs in which the sampling anomaly event was located during a predetermined historical period; determining the probability of the sampling anomaly event occurring in the battery management system and / or the target battery pack based on the sampling anomaly location data corresponding to the plurality of maintenance data packets; and generating combined prompt information based on the location result of the sampling anomaly event and the probability of the anomaly.

[0012] According to another aspect of the present invention, a battery sampling anomaly identification device is provided, comprising: a sampling module, configured to use a battery management system to sample multiple target battery cells included in a target battery pack to obtain measured voltage data corresponding to each of the multiple target battery cells; an identification module, configured to use a target neural network model to identify the measured voltage data corresponding to each of the multiple target battery cells and determine the anomaly identification result of the target battery pack, wherein the target neural network model is trained using a training set, the training set being generated based on voltage feature data corresponding to multiple reference battery packs, the multiple reference battery packs being in different states; a trend determination module, configured to, when the anomaly identification result indicates the presence of a sampling voltage anomaly, determine the voltage time-series trend corresponding to each of the multiple target battery cells based on the measured voltage data corresponding to each of the multiple target battery cells; and a location module, configured to, based on the voltage time-series trend corresponding to each of the multiple target battery cells, locate the sampling anomaly event occurring in the battery management system and / or the target battery pack.

[0013] According to another aspect of the present invention, a non-volatile storage medium is provided, the non-volatile storage medium storing a plurality of instructions adapted for loading by a processor and executing any one of the battery sampling anomaly identification methods described herein.

[0014] According to another aspect of the present invention, an electronic device is provided, comprising: one or more processors and a memory, the memory being used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors cause the one or more processors to implement any one of the battery sampling anomaly identification methods.

[0015] In this embodiment of the invention, a battery management system is used to sample multiple target battery cells included in a target battery pack to obtain measured voltage data corresponding to each of the multiple target battery cells. A target neural network model is used to identify the measured voltage data corresponding to each of the multiple target battery cells to determine the anomaly identification result of the target battery pack. The target neural network model is trained using a training set generated based on voltage feature data corresponding to multiple reference battery packs in different states. When the anomaly identification result indicates the presence of a sampling voltage anomaly, the voltage time-series trend corresponding to each of the multiple target battery cells is determined based on the measured voltage data corresponding to each of the multiple target battery cells. Based on the voltage time-series trend corresponding to each of the multiple target battery cells, the sampling anomaly event is located in the battery management system and / or the target battery pack. This achieves the goal of reducing the workload of sampling anomaly investigation, realizes the technical effect of locating the location of sampling anomalies, and thus solves the technical problem of insufficient sampling anomaly identification and location capabilities in related technologies. Attached Figure Description

[0016] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and, together with their description, serve to explain the invention and do not constitute an undue limitation thereof. In the drawings:

[0017] Figure 1 This is a flowchart of an optional battery sampling anomaly identification method provided according to an embodiment of the present invention;

[0018] Figure 2 This is a schematic diagram of an optional battery sampling anomaly identification method provided according to an embodiment of the present invention;

[0019] Figure 3 This is an algorithm structure diagram of an optional battery sampling anomaly identification method provided according to an embodiment of the present invention;

[0020] Figure 4 This is a schematic diagram of a sliding window for an optional battery sampling anomaly identification method provided according to an embodiment of the present invention;

[0021] Figure 5This is a schematic diagram of an optional battery sampling anomaly identification method provided according to an embodiment of the present invention;

[0022] Figure 6 This is a schematic diagram of a sampling anomaly in an optional battery sampling anomaly identification method provided according to an embodiment of the present invention;

[0023] Figure 7 This is a schematic diagram of a sampling anomaly in another optional battery sampling anomaly identification method provided by an embodiment of the present invention;

[0024] Figure 8 This is a schematic diagram of a sampling anomaly in another optional battery sampling anomaly identification method provided according to an embodiment of the present invention;

[0025] Figure 9 This is a schematic diagram of an optional battery sampling anomaly identification device provided according to an embodiment of the present invention. Detailed Implementation

[0026] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0027] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0028] The battery is the sole power source for pure electric vehicles, and the quality of battery technology directly affects the overall performance of the electric vehicle, including range and power. Therefore, battery safety is also a crucial factor. During vehicle use, to ensure the healthy operation of the battery, the power battery management system monitors the voltage of individual battery cells in real time to prevent overcharging and over-discharging. According to the battery management control strategy, if battery cell sampling is lost or abnormal (voltage too low, too high, etc.) during vehicle operation or charging, certain fault handling measures will be taken, such as limiting the battery pack output power or stopping charging. The battery pack's individual cell sampling module involves individual battery cells, modules, sampling chips, wiring harnesses, and their connectors; failure of any of these modules will lead to the failure of the sampling function module.

[0029] A power battery pack is made up of many individual battery cells connected in series and parallel. During the use of the power battery, the BMS needs to monitor the voltage information of each individual battery cell in real time for SOC calculation, safety monitoring, fault diagnosis, etc. When the individual cell voltage sampling function is lost, such as the sampled voltage is 0, the sampled voltage is higher or lower than the actual voltage, or the voltage does not match the actual voltage, this state of loss of specified function is usually called "failure" or "fault". After the functional abnormality occurs, it is necessary to locate and troubleshoot the abnormality to ensure the normal use of the power battery pack.

[0030] Power batteries are protected against water and dust. However, disassembly and troubleshooting after battery pack assembly can cause physical damage to the batteries, such as damage to the battery casing, electrolyte leakage, or broken connectors, leading to malfunction or safety issues. If the connection between the positive and negative terminals is accidentally short-circuited during maintenance, it can cause a large current flow, resulting in serious safety problems. Generally, it is best to avoid disassembling and inspecting assembled battery packs unless absolutely necessary.

[0031] Since the Battery Management System (BMS) and the battery pack are connected via sampling lines, and the BMS is often located within the control unit, troubleshooting is much easier than disassembling the entire battery pack. If the sampling anomaly can be located from the BMS, a significant amount of disassembly work can be avoided. Current technologies can only confirm that a sampling anomaly has occurred, but cannot identify its specific location, necessitating complete disassembly, resulting in a large amount of repair work and significant difficulty in reassembling the battery pack.

[0032] To address the aforementioned problems, this invention provides a method embodiment for identifying abnormal battery sampling. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0033] Figure 1 This is a flowchart of a battery sampling anomaly identification method according to an embodiment of the present invention, such as... Figure 1 As shown, the method includes the following steps:

[0034] Step S102: Using a battery management system, sample multiple target battery cells included in the target battery pack to obtain the measured voltage data corresponding to each of the multiple target battery cells.

[0035] As we understand it, a Battery Management System (BMS) is used to detect parameters such as voltage, current, and temperature of all target battery cells in a target battery pack. It can monitor the real-time operating status of the target battery pack, including battery capacity and remaining charge. By sampling multiple target battery cells, the BMS can provide protection against overcharging, over-discharging, overcurrent, and overtemperature to prevent battery damage or accidents. If the sampling function malfunctions instead of the battery itself being faulty, the BMS may misjudge based on measured voltage data. For example, it might determine that a battery cell with a low voltage due to a sampling malfunction is over-discharged, and the BMS will cut off power output to the target battery pack, resulting in power loss.

[0036] Step S104: Using a target neural network model, the measured voltage data corresponding to multiple target battery cells are identified to determine the anomaly identification result of the target battery pack. The target neural network model is trained using a training set, which is generated based on the voltage feature data corresponding to multiple reference battery packs, which are in different states.

[0037] It is understandable that by using a trained target neural network model to identify the measured voltage data corresponding to multiple target battery cells, anomaly identification results can be obtained. The training set of the aforementioned target neural network model is obtained based on the voltage characteristic data of multiple reference battery packs. Since the sampling function results in many types of anomalies, and the voltage manifestations of different types of sampling anomalies are different, in order to improve the target neural network model's ability to identify sampling anomalies, reference battery packs in both normal and different anomaly states are selected. This enriches the training set, thereby enhancing the target neural network model's identification ability and leading to more accurate anomaly identification results.

[0038] Optionally, such as Figure 2 As shown, the target neural model described above can contain stacked convolutional kernels and pooling layers. Multiple fully connected layers are added to the back end of the network. A convolutional kernel is a filter used to extract local features from voltage feature data; it extracts features by sliding across the input data and performing a convolution operation. Pooling layers are used to reduce the size of the feature map, thereby reducing computational cost and the number of parameters.

[0039] The specific model structure can be of various types, such as Figure 3 The diagram shows a preferred model setup. The model sequentially passes through 16 convolutional kernels of size 3*d, 32 pooling layers with a 2x1 pooling window, and 64 convolutional kernels of size 3*1. This process is repeated with 32 pooling layers with a 2x1 pooling window and 64 more convolutional kernels of size 3*1. Finally, the model outputs the anomaly detection result after passing through two fully connected layers. For the 16 convolutional kernels of size 3*d, 3 represents the width and height of each kernel in the input data, and d represents the depth or number of channels in the input data. The same logic applies to the 64 convolutional kernels of size 3*1. The 2x1 pooling layers mean that each pooling window is 2x1, meaning that each pooling operation takes the maximum or average value from the two rows as the output. The 31 2x1 pooling layers indicate that there are 31 such pooling layers in the model, and the input to each pooling layer is the feature map from the previous layer. Each neuron in the fully connected layer is connected to all neurons in the previous layer, and each connection has a corresponding weight. Therefore, a fully connected layer can connect each feature of the input data to each neuron in the next layer, and adjust the influence of the input data through weights. By stacking multiple fully connected layers, neural networks can learn more complex and abstract feature representations, and use these features to perform tasks such as classification and regression.

[0040] In an optional embodiment, the method further includes: acquiring voltage feature data of any battery pack among a plurality of reference battery packs; augmenting the voltage feature data of any battery pack to obtain augmented samples corresponding to any battery pack; using the plurality of reference battery packs as arbitrary battery packs to obtain augmented samples corresponding to the plurality of reference battery packs respectively; and labeling the augmented samples corresponding to the plurality of reference battery packs to obtain a training set.

[0041] It is understandable that due to the characteristics of sampling anomalies, such as poor contact leading to abnormalities, recovery may be due to vibration or shaking, resulting in a relatively small amount of data collected. Using voltage characteristic data from normal states as positive samples and voltage characteristic data from different abnormal states as negative samples presents a significant data gap for machine learning models, leading to insufficient ability to identify sampling anomalies. Therefore, expanding any battery pack from multiple reference battery packs yields expanded samples. After standardizing these expanded samples, a training set for training the target neural network model can be obtained.

[0042] Optionally, the augmented samples corresponding to multiple reference battery packs are labeled. Augmented samples containing sampling anomalies are labeled as 1, and samples without sampling anomalies are labeled as 0. All augmented samples are divided into training, validation, and test sets at 60%, 20%, and 20% respectively. The training set is used to train the target neural network model. The model learns and optimizes using the training set data through repeated iterations to better fit the training data. The validation set is used to adjust model hyperparameters and select a model. During training, by evaluating the model's performance on the validation set, it is possible to determine whether the model is overfitting or underfitting and make corresponding adjustments. For example, by changing hyperparameters such as the number of layers, nodes, and learning rate, and then evaluating the model's performance on the validation set, the optimal hyperparameter configuration can be selected. The test set is used to evaluate the final model performance. The test set is used after the training and validation phases to simulate the model's performance in real-world applications. By evaluating on the test set, the model's generalization ability can be tested, i.e., the model's ability to process unseen data, comparing the merits of different models, or evaluating and improving the model.

[0043] Optionally, the voltage characteristic data corresponding to any battery pack in multiple reference battery packs can be obtained by data cleaning and feature extraction. For example, the unprocessed data of any battery pack can be used as the initial voltage data. The sampling interval of the initial voltage data is between 10 and 30 seconds. Packet loss during data transmission is cleaned, including dividing the data into segments according to the time series (segments are formed if there is no data for more than t minutes, usually t is 5 minutes), removing noise, and filling in a small number of missing values. Key features such as the mean voltage, variance, and voltage of adjacent cells of the battery pack are extracted to generate the voltage characteristic data of the aforementioned arbitrary battery pack.

[0044] In one optional embodiment, the voltage characteristic data of any battery pack is expanded to obtain an expanded sample corresponding to any battery pack. This includes: based on a predetermined sliding step size and a predetermined window timing length, the voltage characteristic data of any battery pack is truncated multiple times using a sliding window to obtain multiple sets of window data that overlap; if the timing length of any set of data in the multiple sets of window data is less than the predetermined window timing length, the set of data is spliced ​​together so that the timing length of any set of data is equal to the predetermined window timing length.

[0045] It is understandable that a sliding window approach can be used to repeatedly extract voltage characteristic data from any battery pack. By partially overlapping the extraction windows, multiple sets of window data can be obtained. To ensure that the sample lengths are consistent, if the length of any set of data in the multiple sets of window data is less than the predetermined window time series length (generally the end of the data), these sets of data can be concatenated so that the time series length of any set of data equals the predetermined window time series length.

[0046] Optionally, such as Figure 4 As shown, with a predetermined sliding step size of s and a predetermined window time length of n, it can be seen that some windows cover sampling anomalies in adjacent battery cells, which can be marked as 1. Other windows may or may not include sampling anomalies. Using this method, multiple positive and negative samples can be obtained. Figure 5 As shown, when the window is not fully captured, such as capturing a partial window of type k and another partial window of type nk, the two windows can be spliced ​​together so that the sample length is the predetermined window time series length n.

[0047] Step S106: If the anomaly identification result indicates that there is an abnormal sampling voltage, determine the voltage timing trend of each of the multiple target battery cells based on the measured voltage data corresponding to each of the multiple target battery cells.

[0048] It is understandable that when using a target neural network for anomaly identification and determining that there is an abnormal sampling voltage, the sampling is affected by the state of the battery pack side and the state of the battery management system side. After identifying the abnormal sampling voltage, it is also necessary to locate the abnormal sampling. Based on the measured voltage data corresponding to multiple target battery cells, the voltage timing trend corresponding to multiple target battery cells is determined.

[0049] Step S108: Based on the voltage timing trends of multiple target battery cells, locate the sampling anomaly event that occurred in the battery management system and / or the target battery pack.

[0050] It is understandable that the battery management system samples multiple target battery cells in real time, and each target battery cell has a voltage change trend over time, i.e., a corresponding voltage time-series trend. Based on the voltage time-series trends corresponding to multiple target battery cells, the location of sampling anomalies can be located in the battery management system and / or the target battery pack.

[0051] In one optional embodiment, based on the voltage timing trends corresponding to multiple target battery cells, the location of a sampling anomaly event occurring in the battery management system and / or the target battery pack includes: when there are two adjacent battery cells among the multiple target battery cells, and the voltage timing trends corresponding to the two adjacent battery cells are symmetrically distributed about the mean change trend, and the voltage timing trend of one of the two adjacent battery cells is higher than the mean change trend, while the voltage timing trend of the other battery cell is lower than the mean change trend, the location of the sampling anomaly event occurring in the target battery pack is determined, wherein the mean change trend is the average voltage trend of the other cells among the multiple target battery cells excluding the two adjacent cells.

[0052] Under normal circumstances, the voltages of all target battery cells in a battery pack are relatively balanced, and the voltage timing trends of each cell tend to be consistent. When a battery sampling anomaly occurs, compared to anomalies in the battery cells themselves (such as over-discharge), the voltage timing trend caused by the sampling anomaly has a identifiable specific trend. If there are two adjacent battery cells among multiple target cells, and the voltage timing trends of these two adjacent cells are symmetrically distributed about the mean, and one cell's voltage timing trend is higher than the mean while the other's is lower than the mean—that is, the two adjacent cells are one high and one low, and their voltage timing trends are symmetrically distributed about the mean—then the sampling anomaly can be located within the target battery pack.

[0053] Optionally, in the above cases, the sampling anomaly event can be of various types, such as Figure 6 As shown, on one side of the battery pack, the shared sampling line between two adjacent cells has increased contact impedance due to adhesive residue at the terminals or abnormal connections. Another possibility is a welding failure or missing weld between adjacent cells, resulting in a weak series connection. This leads to a difference in voltage levels between adjacent cells, symmetrically distributed along the mean. The reason for this is that the cell with the higher voltage... Figure 6 In the sampled line, the voltage value of U1 is the difference between V2 and V1. Due to abnormal sampling line or loose series connection, the contact resistance increases abnormally, which leads to an increase in the voltage drop of U1 and a larger V2. Since the two connected cells share a common line, the voltage of U2 is obtained from the difference between sampling line V3 and V2, which will cause the voltage of the adjacent cell U2 to be lower.

[0054] In one optional embodiment, based on the voltage timing trends corresponding to multiple target battery cells, the location of a sampling anomaly event occurring in the battery management system and / or the target battery pack includes: if there are three adjacent battery cells among the multiple target battery cells, namely the first cell, the second cell, and the third cell, and the voltage timing trends corresponding to the first cell and the third cell are the same and higher than the average change trend, while the voltage timing trend of the second cell is lower than the average change trend, the location of the sampling anomaly event occurring in the battery management system is determined. The average change trend is the average voltage trend of the other cells among the multiple target battery cells excluding the three adjacent cells.

[0055] It is understandable that among multiple target battery cells, there are three adjacent battery cells whose voltage timing trends are characterized by the following: the voltage timing trends of the first and third cells are the same and higher than the average trend, while the voltage timing trend of the second cell is lower than the average trend. Based on the above characteristics, the sampling anomaly event can be located in the battery management system.

[0056] It should be noted that in the above situations, such as Figure 7 The sampling anomaly event shown could be due to excessive leakage current in the Zener diode of the second cell within the battery management system. The Zener diode limits the maximum output current of the battery cell, preventing overload or short circuits and protecting the battery cell and the entire system. Excessive leakage current in the second cell's Zener diode leads to increased voltage division on the shared acquisition line between the first and second cells, as well as increased voltage division on the shared acquisition line between the third and second cells. This results in the voltage timing trends for the first and third cells being identical and higher than the average trend, while the timing trend for the second cell is lower than the average trend. Furthermore, the first voltage difference between the first and third cells exceeding the average trend is twice the second voltage difference between the second cell and the second cell.

[0057] In one optional embodiment, based on the voltage timing trends corresponding to multiple target battery cells, the sampling anomaly event is located in the battery management system and / or the target battery pack. This includes: when there are two adjacent battery cells among the multiple target battery cells, and the voltage timing trends corresponding to the two adjacent battery cells are simultaneously higher or lower than the average change trend, the sampling anomaly event is located in the target battery pack and the battery management system. The average change trend is the average voltage trend of the other cells among the multiple target battery cells, excluding the two adjacent cells.

[0058] It is understandable that there are multiple sampling anomalies. When multiple anomalies coexist, the voltage timing trend is characterized by two adjacent battery cells, where the voltage timing trends of the two adjacent battery cells are simultaneously higher or lower than the average change trend. This can help pinpoint the sampling anomaly to the target battery pack and battery management system.

[0059] It should be noted that, in the above situations, such as Figure 8 As shown, under abnormal conditions on both the battery management system side and the battery pack side, multiple abnormalities may overlap, resulting in adjacent cell voltages being either high or low, possibly both at or below the average voltage value, rather than in a symmetrical form.

[0060] In an optional embodiment, after locating the sampling anomaly event occurring in the battery management system and / or the target battery pack based on the voltage timing trends corresponding to multiple target battery cells, the method further includes: determining multiple maintenance battery packs matching the target battery pack type, and sampling anomaly location data corresponding to multiple maintenance data packets, wherein the multiple maintenance battery packs are battery packs whose sampling anomaly events were located in a predetermined historical period; determining the anomaly probability of the sampling anomaly event occurring in the battery management system and / or the target battery pack based on the sampling anomaly location data corresponding to the multiple maintenance data packets; and generating combined prompt information based on the location result of the sampling anomaly event and the anomaly probability.

[0061] It is understandable that, in order to further improve the accuracy of the prompt information, based on the sampling anomaly location data corresponding to multiple battery packs of the same type, the location where the sampling anomaly of that type of battery pack has occurred can be indicated, and the anomaly probability in the battery management system or the target battery pack can be obtained. After combining the location result with the anomaly probability, combined prompt information is obtained. The purpose of locating the anomaly event is to provide relevant technicians with repair suggestions, facilitate the investigation of the fault location, and avoid unnecessary disassembly of the battery pack.

[0062] Through the above steps S102, the battery management system samples multiple target battery cells included in the target battery pack to obtain the measured voltage data corresponding to each target battery cell. In step S104, a target neural network model is used to identify the measured voltage data corresponding to each target battery cell, determining the anomaly identification result of the target battery pack. The target neural network model is trained using a training set generated based on the voltage feature data corresponding to multiple reference battery packs in different states. In step S106, if the anomaly identification result indicates an abnormal sampling voltage, the voltage time-series trend corresponding to each target battery cell is determined based on the measured voltage data. In step S108, based on the voltage time-series trend corresponding to each target battery cell, the sampling anomaly event is located in the battery management system and / or the target battery pack. This reduces the workload of sampling anomaly investigation and achieves the technical effect of locating the location of sampling anomalies, thus solving the technical problem of insufficient sampling anomaly identification and location capabilities in related technologies.

[0063] Based on the above embodiments and optional embodiments, the present invention proposes an optional implementation method for use in electric vehicles with lithium-ion batteries. The battery management system in the control unit of the vehicle is connected to the power battery pack through communication lines and sampling lines to collect data on individual battery cells in real time and obtain measured voltage data.

[0064] To identify sampling anomalies, the target neural network model is first required. This model is generated by acquiring initial voltage data from different reference battery packs (both normal and abnormal), extracting data sentiment and features, and generating voltage feature data. To ensure a relative balance between positive and negative samples for normal and abnormal states, a sliding window approach is used to augment the samples. After labeling the augmented samples, the training, validation, and test sets are divided into 60%, 20%, and 20% sets, respectively. The target neural network model is trained using the training set, the model parameters are adjusted during training using the validation set, and finally, the model performance is validated using the test set, resulting in the target neural network model.

[0065] By inputting the measured voltage data into the target neural network model for identification, the anomaly identification results of the battery pack can be obtained. If an anomaly in the sampled voltage of the battery pack is determined, the voltage time-series trend of each individual battery cell can be identified based on the measured voltage data. By identifying trend characteristics, such as a high-low pattern between two adjacent battery cells that shows symmetry along the average voltage trend, the sampling anomaly can be located within the battery pack. Similarly, if three adjacent battery cells exhibit a high-low-high trend, and the anomaly trends of the first and third cells are identical, the sampling anomaly can be located within the battery management system.

[0066] The above method can locate sampling anomalies in the battery pack. By combining this with the type of battery pack being repaired, the probability of anomalies occurring in that type of battery pack can be determined. Combined prompts can then be provided to relevant technicians to assist in troubleshooting sampling anomalies. This effectively improves the ability to identify sampling anomalies and reduces the workload of disassembling batteries to troubleshoot sampling anomalies.

[0067] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0068] This embodiment also provides a battery sampling anomaly identification device, which is used to implement the above embodiments and preferred embodiments, and will not be repeated as already described. As used below, the terms "module" and "device" can refer to a combination of software and / or hardware that implements a predetermined function. Although the device described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0069] According to embodiments of the present invention, an apparatus embodiment for implementing a battery sampling anomaly identification method is also provided. Figure 9 This is a schematic diagram of a battery sampling anomaly identification device according to an embodiment of the present invention, such as... Figure 9 As shown, the above-mentioned battery sampling anomaly identification device includes: a sampling module 902, an identification module 904, a trend determination module 906, and a positioning module 908. The device will be described below.

[0070] The sampling module 902 is used to sample multiple target battery cells included in the target battery pack using the battery management system, and obtain the measured voltage data corresponding to each of the multiple target battery cells.

[0071] The identification module 904, connected to the sampling module 902, is used to identify the measured voltage data corresponding to multiple target battery cells using a target neural network model, and to determine the abnormal identification result of the target battery pack. The target neural network model is trained using a training set, which is generated based on the voltage feature data corresponding to multiple reference battery packs, which are in different states.

[0072] The trend determination module 906 is connected to the identification module 904 and is used to determine the voltage time sequence trend of multiple target battery cells based on the measured voltage data corresponding to multiple target battery cells when the abnormal identification result indicates that there is an abnormal sampling voltage.

[0073] The positioning module 908, connected to the trend determination module 906, is used to locate the sampling abnormal event that occurred in the battery management system and / or the target battery pack based on the voltage timing trend corresponding to multiple target battery cells.

[0074] In a battery sampling anomaly identification device provided in this embodiment of the invention, a sampling module 902 is used to sample multiple target battery cells included in a target battery pack using a battery management system to obtain measured voltage data corresponding to each of the multiple target battery cells; an identification module 904, connected to the sampling module 902, is used to identify the measured voltage data corresponding to each of the multiple target battery cells using a target neural network model to determine the anomaly identification result of the target battery pack, wherein the target neural network model is trained using a training set, which is generated based on voltage feature data corresponding to multiple reference battery packs in different states; a trend determination module 906, connected to the identification module 904, is used to determine the voltage time-series trend of each of the multiple target battery cells based on the measured voltage data corresponding to each of the multiple target battery cells when the anomaly identification result indicates that there is a sampling voltage anomaly; and a positioning module 908, connected to the trend determination module 906, is used to locate the sampling anomaly event occurring in the battery management system and / or the target battery pack based on the voltage time-series trend of each of the multiple target battery cells. This achieves the goal of reducing the workload of investigating sampling anomalies, realizes the technical effect of locating the location of sampling anomalies, and thus solves the technical problem of insufficient sampling anomaly identification and location capabilities in related technologies.

[0075] It should be noted that the above modules can be implemented by software or hardware. For example, for the latter, it can be implemented in the following ways: the above modules can be located in the same processor; or the above modules can be located in different processors in any combination.

[0076] It should be noted that the sampling module 902, identification module 904, trend determination module 906, and positioning module 908 mentioned above correspond to steps S102 to S108 in the embodiments. The instances and application scenarios implemented by the above modules and corresponding steps are the same, but are not limited to the content disclosed in the above embodiments. It should be noted that the above modules, as part of the device, can run in a computer terminal.

[0077] It should be noted that the optional or preferred implementation methods of this embodiment can be found in the relevant descriptions in the embodiments, and will not be repeated here.

[0078] The aforementioned battery sampling anomaly identification device may also include a processor and a memory. The sampling module 902, the identification module 904, the trend determination module 906, the positioning module 908, etc., are all stored in the memory as program units, and the processor executes the aforementioned program units stored in the memory to realize the corresponding functions.

[0079] The processor contains a core that retrieves the corresponding program unit from memory. One or more cores may be configured. Memory may include non-persistent memory in computer-readable media, random access memory (RAM), and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory includes at least one memory chip.

[0080] This invention provides a non-volatile storage medium storing a program that, when executed by a processor, implements a battery sampling anomaly identification method.

[0081] This invention provides an electronic device including a processor, a memory, and a program stored in the memory and executable on the processor. When the processor executes the program, it performs the following steps: using a battery management system to sample multiple target battery cells in a target battery pack, obtaining measured voltage data corresponding to each target battery cell; using a target neural network model to identify the measured voltage data corresponding to each target battery cell, determining anomaly identification results for the target battery pack, wherein the target neural network model is trained using a training set generated based on voltage feature data corresponding to multiple reference battery packs in different states; when the anomaly identification results indicate an abnormal sampling voltage, determining the voltage timing trend corresponding to each target battery cell based on the measured voltage data; and locating the sampling anomaly event occurring in the battery management system and / or the target battery pack based on the voltage timing trend corresponding to each target battery cell. The device in this document can be a server, PC, etc.

[0082] This invention also provides a computer program product, which, when executed on a data processing device, is suitable for executing an initialization program with the following steps: using a battery management system to sample multiple target battery cells included in a target battery pack, obtaining measured voltage data corresponding to each of the multiple target battery cells; using a target neural network model to identify the measured voltage data corresponding to each of the multiple target battery cells, determining the anomaly identification result of the target battery pack, wherein the target neural network model is trained using a training set, which is generated based on voltage feature data corresponding to multiple reference battery packs, which are in different states; when the anomaly identification result indicates the presence of a sampling voltage anomaly, determining the voltage timing trend corresponding to each of the multiple target battery cells based on the measured voltage data corresponding to each of the multiple target battery cells; and locating the sampling anomaly event occurring in the battery management system and / or the target battery pack based on the voltage timing trend corresponding to each of the multiple target battery cells.

[0083] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0084] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0085] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0086] These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable apparatus for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0087] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0088] Memory may include non-persistent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0089] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0090] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0091] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0092] The above are merely embodiments of the present invention and are not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principle of the present invention should be included within the scope of the claims of the present invention.

Claims

1. A battery sampling abnormality identification method, characterized by, include: A battery management system is used to sample multiple target battery cells included in the target battery pack to obtain the measured voltage data corresponding to each of the multiple target battery cells. A target neural network model is used to identify the measured voltage data corresponding to the multiple target battery cells, and to determine the anomaly identification result of the target battery pack. The target neural network model is trained using a training set, which is generated based on the voltage feature data corresponding to multiple reference battery packs, which are in different states. When the anomaly identification result indicates that there is an abnormal sampling voltage, the voltage timing trend of each of the multiple target battery cells is determined based on the measured voltage data corresponding to each of the multiple target battery cells. Based on the voltage timing trends corresponding to the multiple target battery cells, the sampling anomaly event is located in the battery management system and / or the target battery pack. The step of locating the sampling anomaly event occurring in the battery management system and / or the target battery pack based on the voltage time-series trends corresponding to the plurality of target battery cells includes: if there are two adjacent battery cells among the plurality of target battery cells, and the voltage time-series trends corresponding to the two adjacent battery cells are symmetrically distributed about the mean change trend, and the voltage time-series trend of one of the two adjacent battery cells is higher than the mean change trend, while the voltage time-series trend of the other battery cell is lower than the mean change trend, then the sampling anomaly event is located in the target battery pack. The mean change trend is the average voltage trend of the other battery cells among the plurality of target battery cells excluding the two adjacent battery cells. If there are three adjacent battery cells among the plurality of target battery cells, sequentially named the first battery cell, the second battery cell, the third battery cell, the fourth battery cell, the fifth battery cell, the sixth battery cell, the seventh battery cell, the eleventh ... In the case of two individual cells and a third individual cell, where the voltage timing trends of the first and third individual cells are the same and higher than the average change trend, and the voltage timing trend of the second individual cell is lower than the average change trend, the sampling anomaly event is located in the battery management system. The average change trend is the average voltage trend of the individual cells other than the three adjacent cells in the plurality of target battery cells. In the case of two adjacent battery cells in the plurality of target battery cells, where the voltage timing trends of the two adjacent battery cells are simultaneously higher or lower than the average change trend, the sampling anomaly event is located in the target battery pack and the battery management system. The average change trend is the average voltage trend of the individual cells other than the two adjacent cells in the plurality of target battery cells.

2. The method of claim 1, wherein, The method further includes: For any battery pack among the plurality of reference battery packs, obtain the voltage characteristic data of that battery pack; The voltage characteristic data of any battery pack is augmented to obtain the augmented sample corresponding to the arbitrary battery pack; By using the plurality of reference battery packs as the arbitrary battery pack, the augmented samples corresponding to the plurality of reference battery packs are obtained respectively; The training set is obtained by labeling the expanded samples corresponding to the multiple reference battery packs.

3. The method according to claim 2, characterized in that, The step of augmenting the voltage characteristic data of the arbitrary battery pack to obtain the augmented sample corresponding to the arbitrary battery pack includes: Based on a predetermined sliding step size and a predetermined window timing length, the voltage characteristic data of any battery pack is repeatedly truncated using a sliding window to obtain multiple sets of window data with overlap. If the time sequence length of any one of the multiple sets of window data is less than the predetermined window time sequence length, the data of any one set is concatenated so that the time sequence length of the data of any one set is equal to the predetermined window time sequence length.

4. The method according to any one of claims 1 to 3, characterized in that, The method further includes, after locating the sampling anomaly event occurring after the battery management system and / or the target battery pack, based on the voltage timing trends corresponding to the plurality of target battery cells, the method further includes: Identify multiple repair battery packs that match the target battery pack type, and sampling anomaly location data corresponding to multiple repair data packets, wherein the multiple repair battery packs are battery packs whose sampling anomaly events were located in a predetermined historical period; Based on the sampling anomaly location data corresponding to the multiple maintenance data packages, the probability of the sampling anomaly event occurring in the battery management system and / or the target battery pack is determined. Based on the location results of the sampled abnormal events and the abnormal probability, a combined prompt message is generated.

5. A battery sampling anomaly identification device, characterized in that, include: The sampling module is used to sample multiple target battery cells included in the target battery pack using the battery management system, and obtain the measured voltage data corresponding to each of the multiple target battery cells. The identification module is used to identify the measured voltage data corresponding to the multiple target battery cells using a target neural network model, and to determine the abnormal identification result of the target battery pack. The target neural network model is trained using a training set, which is generated based on the voltage feature data corresponding to multiple reference battery packs, which are in different states. The trend determination module is used to determine the voltage time sequence trend of each of the multiple target battery cells based on the measured voltage data corresponding to each of the multiple target battery cells when the anomaly identification result indicates that there is an abnormal sampling voltage. The positioning module is used to locate the sampling abnormal event that occurred in the battery management system and / or the target battery pack based on the voltage timing trend corresponding to the multiple target battery cells respectively; The positioning module is further configured to locate the sampling anomaly event occurring in the target battery pack when there are two adjacent battery cells among the plurality of target battery cells, and the voltage time-series trends of the two adjacent battery cells are symmetrically distributed about the mean change trend, and the voltage time-series trend of one of the two adjacent battery cells is higher than the mean change trend, while the voltage time-series trend of the other battery cell is lower than the mean change trend. The mean change trend is the average voltage trend of the other battery cells among the plurality of target battery cells excluding the two adjacent cells. The module is further configured to locate the sampling anomaly event occurring in the target battery pack when there are three adjacent battery cells among the plurality of target battery cells, namely, a first cell, a second cell, and a third cell, wherein the first cell and the third cell are... If the voltage timing trends of the two target battery cells are the same and higher than the average trend, and the voltage timing trend of the second cell is lower than the average trend, the sampling anomaly is located in the battery management system. The average trend is the average voltage trend of the other cells among the multiple target battery cells, excluding the three adjacent cells. If there are two adjacent battery cells among the multiple target battery cells, and the voltage timing trends of the two adjacent battery cells are simultaneously higher or lower than the average trend, the sampling anomaly is located in both the target battery pack and the battery management system. The average trend is the average voltage trend of the other cells among the multiple target battery cells, excluding the two adjacent cells.

6. A non-volatile storage medium, characterized in that, The non-volatile storage medium stores multiple instructions, which are adapted to be loaded by a processor and executed by the battery sampling anomaly identification method according to any one of claims 1 to 4.

7. An electronic device, characterized in that, include: One or more processors and a memory, the memory being used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors cause the one or more processors to implement the battery sampling anomaly identification method according to any one of claims 1 to 4.

Images