Battery abnormality warning method, device, equipment, readable storage medium and program product

Through the multi-layer preset model, the state characteristic data of the power battery is analyzed, and the problem of low accuracy of battery thermal runaway warning in the prior art is solved, and a more accurate battery abnormal warning is achieved.

CN118722237BActive Publication Date: 2024-12-06ZHEJIANG LEAPENERGY TECH CO LTD
View PDF 1 Cites 0 Cited by

Patent Information

Application Number
CN202411218771.8
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-09-02
Publication Date
2024-12-06
Estimated Expiration
2044-09-02

AI Technical Summary

Technical Problem

The prior art has low accuracy in thermal runaway warning of power batteries, which makes it difficult to warn in a timely manner in case of sudden heat runaway.

Method used

By obtaining the battery status data of the vehicle during the target charging cycle, the battery status characteristic data is extracted using a multi-layer preset model (including system encoder, system decoder and timing encoder, and timing decoder), combining short and long cycles. The charging process data calculates the battery abnormality warning result.

Benefits of technology

It improves the accuracy of battery abnormal warning, can promptly detect short- and long-term abnormalities in battery operation, and reduces hidden dangers such as thermal runaway.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118722237B_ABST
    Figure CN118722237B_ABST
Patent Text Reader

Abstract

The present application relates to a battery abnormality warning method, device, equipment, readable storage medium and program product. The method comprises: first, obtaining the battery status data of a vehicle in a target charging cycle, the target charging cycle includes multiple charging processes, then, inputting the battery status data into a first preset model, obtaining a plurality of first state characteristic data for characterizing the state characteristic data corresponding to each charging process of the vehicle, then, inputting the plurality of first state characteristic data into a second preset model, obtaining the state characteristic data corresponding to multiple charging processes in the target charging cycle of the vehicle, and finally, determining the battery abnormality warning result of the vehicle according to the second state characteristic data. The use of this method can improve the accuracy of battery abnormality warning.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of battery technology, and in particular to a battery abnormality warning method, device, equipment, readable storage medium and program product. Background Art

[0002] With the development of electric vehicles, electric vehicles account for an increasingly higher proportion in the passenger car market. Therefore, the safety of electric vehicles has attracted more attention from the market. Thermal runaway of power batteries in electric vehicles due to internal short circuits is one of the main safety hazards of electric vehicles. Since thermal runaway often occurs suddenly in a short period of time and there are almost no obvious characteristics before it occurs, it is particularly important to provide timely warning of thermal runaway of power batteries.

[0003] At present, the accuracy of early warning of thermal runaway of power batteries is not high. Summary of the invention

[0004] Based on this, it is necessary to provide a more accurate battery abnormality warning method, device, equipment, readable storage medium and program product to address the above technical problems.

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

[0006] Obtaining battery status data of the vehicle within a target charging cycle, where the target charging cycle includes multiple charging processes;

[0007] Inputting the battery state data into a first preset model to obtain a plurality of first state characteristic data; the first state characteristic data is used to characterize the state characteristics corresponding to each charging process of the vehicle within a target charging cycle;

[0008] Inputting a plurality of first state characteristic data into a second preset model to obtain second state characteristic data; the second state characteristic data is used to characterize state characteristics corresponding to multiple charging processes of the vehicle within a target charging cycle;

[0009] The battery abnormality warning result of the vehicle is determined based on the second state characteristic data.

[0010] In one embodiment, the battery status data is input into a first preset model to obtain a plurality of first status characteristic data, including:

[0011] Determining first battery status data corresponding to each charging process of the vehicle according to the battery status data;

[0012] Determine battery excitation data and battery response data from the first battery state data, the battery excitation data is used to characterize external environment data or external input data of the battery during charging, and the battery response data is used to characterize response data of the battery under the influence of the external environment data or external input data;

[0013] The battery excitation data and the battery response data are input into a first preset model to obtain first state characteristic data.

[0014] In one embodiment, the first preset model includes a system encoder and a system decoder, and the battery excitation data and the battery response data are input into the first preset model to obtain the first state characteristic data, including:

[0015] Inputting the battery excitation data and the battery response data into a system encoder to obtain first encoded data;

[0016] Inputting the first coded data and the battery excitation data into a system decoder to obtain first response data;

[0017] Determine first error data according to the first response data and the battery response data;

[0018] First state characteristic data is obtained according to the first coded data and the first error data.

[0019] In one embodiment, the second preset model includes a timing encoder and a timing decoder, and a plurality of first state characteristic data are input into the second preset model to obtain second state characteristic data of the vehicle, including:

[0020] Splicing a plurality of first state feature data according to time to obtain first time series data;

[0021] Inputting the first time series data into a time series encoder to obtain second encoded data;

[0022] Inputting the second encoded data into a timing decoder to obtain second decoded data;

[0023] Determine second error data according to the second decoded data and the first timing data;

[0024] The second coded data and the second error data are used as second state characteristic data.

[0025] In one embodiment, determining a battery abnormality warning result of a vehicle according to the second state characteristic data includes:

[0026] The second state characteristic data is input into a third preset model, and the battery abnormality warning result is determined according to the output result of the third preset model. The third preset model is trained based on the second state characteristic data of sample vehicles, and the sample vehicles include vehicles with battery abnormalities.

[0027] In one embodiment, determining a battery abnormality warning result of a vehicle according to the second state characteristic data includes:

[0028] Determine a battery abnormality warning result according to the second error data in the second state characteristic data and the battery abnormality warning threshold;

[0029] Among them, the battery abnormality warning threshold is calculated based on the mean and standard deviation of the second error data of the sample vehicles, and the sample vehicles do not include vehicles with battery abnormalities.

[0030] In one embodiment, the training process of the first preset model includes:

[0031] Acquire sample battery status data of the sample vehicle, and determine second battery status data corresponding to each charging process of the sample vehicle according to the sample battery status data;

[0032] determining sample battery stimulus data and sample battery response data from the second battery state data;

[0033] Inputting the sample battery excitation data and the sample battery response data into an initial system encoder to obtain first sample encoded data;

[0034] Inputting the first sample encoded data and the sample battery excitation data into the initial system decoder to obtain the first sample decoded data;

[0035] Calculating first sample error data according to the first sample decoded data and the sample battery response data;

[0036] An initial system encoder and an initial system decoder are trained according to the first sample error data to obtain a system encoder and a system decoder.

[0037] In one embodiment, the training process of the second preset model includes:

[0038] Obtaining sample state characteristic data according to the first sample encoding data and the first sample error data;

[0039] The sample state characteristic data is spliced ​​according to time to obtain second time series data;

[0040] Inputting the second time series data into the initial time series encoder to obtain second sample encoded data;

[0041] Inputting the second sample encoded data into the initial timing decoder to obtain second sample decoded data;

[0042] Determine second sample error data according to the second sample decoded data and the second timing data;

[0043] The initial temporal encoder and the initial temporal decoder are trained according to the second sample error data to obtain a temporal encoder and a temporal decoder.

[0044] In a second aspect, the present application also provides a battery abnormality warning device, the device comprising:

[0045] An acquisition module, used to acquire battery status data of a vehicle within a target charging cycle, where the target charging cycle includes multiple charging processes;

[0046] A first input module, used to input the battery state data into a first preset model to obtain a plurality of first state characteristic data; the first state characteristic data is used to characterize the state characteristics corresponding to each charging process of the vehicle within a target charging cycle;

[0047] A second input module is used to input a plurality of first state characteristic data into a second preset model to obtain second state characteristic data; the second state characteristic data is used to characterize state characteristics corresponding to multiple charging processes of the vehicle within a target charging cycle;

[0048] The determination module is used to determine the battery abnormality warning result of the vehicle according to the second state characteristic data.

[0049] In a third aspect, the present application further provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the battery abnormality warning method described in any one of the first aspects above is implemented.

[0050] In a fourth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the battery abnormality warning method described in any one of the first aspects above is implemented.

[0051] In a fifth aspect, the present application further provides a computer program product, including a computer program, which, when executed by a processor, implements the battery abnormality warning method described in any one of the first aspects above.

[0052] The above-mentioned battery abnormality warning method, device, equipment, readable storage medium and program product first obtain the battery status data of the vehicle within the target charging cycle, and the target charging cycle includes multiple charging processes. Then, the battery status data is input into the first preset model to obtain multiple first state characteristic data for characterizing the state characteristic data corresponding to each charging process of the vehicle. Then, the multiple first state characteristic data are input into the second preset model to obtain the state characteristic data corresponding to the multiple charging processes within the target charging cycle of the vehicle. Finally, according to the second state characteristic data, the battery abnormality warning result of the vehicle is determined. In this way, when warning the battery abnormality of the vehicle, the single charging process of the battery is analyzed by the first preset model, and the multiple charging processes of the battery are analyzed by the second preset model. At the same time, the data of the short-cycle and long-cycle charging processes of the battery are considered, so that both the abnormality of the battery in the short-cycle operation and the change of the long cycle can be found, so that the battery abnormality warning result obtained is more accurate. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the drawings required for use in the embodiments of the present application or related technical descriptions will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0054] Figure 1 A schematic diagram of a process of a battery abnormality warning method in one embodiment;

[0055] Figure 2 A flowchart of a battery abnormality warning method in another embodiment;

[0056] Figure 3 A flowchart of a battery abnormality warning method in another embodiment;

[0057] Figure 4 A flowchart of a battery abnormality warning method in another embodiment;

[0058] Figure 5 is a schematic diagram of the structure of a first preset model in an embodiment;

[0059] Figure 6 is a schematic diagram of the structure of a gated recurrent unit in one embodiment;

[0060] Figure 7 A flowchart of a battery abnormality warning method in another embodiment;

[0061] Figure 8A flowchart of a battery abnormality warning method in another embodiment;

[0062] Fig. 9 is a schematic structural diagram of a second preset model in another embodiment;

[0063] Fig.10 A flowchart of a battery abnormality warning method in another embodiment;

[0064] Fig.11 A structural block diagram of a battery abnormality warning device in one embodiment;

[0065] Fig.12 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0066] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0067] In an exemplary embodiment, Figure 1 As shown, a battery abnormality warning method is provided, and the method is applied to a terminal as an example for explanation. It can be understood that the method can also be applied to a server, and can also be applied to a system including a terminal and a server, and is implemented through the interaction between the terminal and the server, including the following steps:

[0068] Step 101, obtaining battery status data of a vehicle within a target charging cycle.

[0069] Among them, the target charging cycle includes multiple charging processes, and optionally, the target charging cycle can be the entire life cycle of the vehicle, and the charging process can include various charging conditions. The battery status data can include data of the following dimensions, such as the total voltage of the power battery pack, the voltage of each cell of the power battery pack, the average cell voltage of the power battery pack, the highest cell voltage of the power battery pack, the lowest cell voltage of the power battery pack, the total current of the power battery pack, the real-time input ampere-hour of this charging, the real-time SOC (State of Charge) of the power battery, the temperature of each battery probe, the real-time ambient temperature, etc. Among them, the charging conditions include fast charging, slow charging, variable current charging, constant current charging or constant voltage charging, etc., as long as it is a charging condition. Optionally, after obtaining the battery status data of all charging conditions in the entire life cycle of the vehicle, the data is cleaned and normalized, which can include clearing the data rows containing null values ​​in the data, and then normalizing the data of each dimension column to a standard normal distribution. Optionally, the vehicle can be one vehicle or multiple vehicles, that is, the battery status data can include the battery status data of one vehicle or multiple vehicles.

[0070] Step 102: Input the battery status data into a first preset model to obtain a plurality of first status characteristic data.

[0071] Among them, the first state characteristic data is used to characterize the state characteristics corresponding to each charging process of the vehicle within the target charging cycle. Optionally, the first preset model can be trained based on the battery state data of the sample vehicle. If the battery state data is the data of one vehicle, multiple first state characteristic data of each charging process of the vehicle can be obtained through the first preset model. If the battery state data is the data of multiple vehicles, multiple first state characteristic data of each charging process of each vehicle can be obtained through the first preset model, so as to analyze the short-term battery data of each vehicle.

[0072] Step 103: input a plurality of first state characteristic data into a second preset model to obtain second state characteristic data.

[0073] Among them, the second state characteristic data is used to characterize the state characteristics corresponding to multiple charging processes of the vehicle within the target charging cycle. Optionally, the first preset model can be trained based on the battery state data of the sample vehicle. If the battery state data is the data of one vehicle, the multiple first state characteristic data of each charging process of the vehicle are spliced ​​according to time to obtain the time series data of all charging processes of the vehicle within the target charging cycle, and input into the second preset model to obtain the second state characteristic data. If the battery state data is the data of multiple vehicles, the multiple first state characteristic data of each charging process of each vehicle are spliced ​​according to time to obtain the time series data of all charging processes of each vehicle within the target charging cycle, and input into the second preset model to obtain the second state characteristic data of each vehicle.

[0074] Step 104, determining a battery abnormality warning result of the vehicle according to the second state characteristic data.

[0075] Among them, the battery abnormality warning result is used to characterize the abnormality of the battery, such as thermal runaway abnormality or short circuit abnormality. Optionally, the battery abnormality warning result can be a warning score of thermal runaway of the battery of the vehicle. According to the battery abnormality warning result, it can be determined whether the battery of the vehicle may have thermal runaway abnormality. According to the second state characteristic data, the battery abnormality warning result of each vehicle can be calculated.

[0076] In the above embodiment, first, the battery status data of the vehicle in the target charging cycle is obtained, and the target charging cycle includes multiple charging processes. Then, the battery status data is input into the first preset model to obtain multiple first state characteristic data for characterizing the state characteristic data corresponding to each charging process of the vehicle. Then, the multiple first state characteristic data are input into the second preset model to obtain the state characteristic data corresponding to the multiple charging processes in the target charging cycle of the vehicle. Finally, according to the second state characteristic data, the battery abnormality warning result of the vehicle is determined. In this way, when warning the battery abnormality of the vehicle, the single charging process of the battery is analyzed by the first preset model, and the multiple charging processes of the battery are analyzed by the second preset model. At the same time, the data of the short-cycle and long-cycle charging processes of the battery are considered, so that both the abnormality of the battery in the short-cycle operation and the change of the long cycle can be found, so that the battery abnormality warning result obtained is more accurate.

[0077] In one embodiment, the above step 102 inputs the battery status data into the first preset model to obtain a plurality of first status characteristic data, such as Figure 2 As shown, including:

[0078] Step 201: Determine first battery status data corresponding to each charging process of the vehicle according to the battery status data.

[0079] The battery status data is divided to determine the first battery status data corresponding to each charging process of a single vehicle.

[0080] Step 202, determining battery excitation data and battery response data from the first battery status data.

[0081] Among them, the battery excitation data is used to characterize the external environment data or external input data of the battery during the charging process, and the battery response data is used to characterize the response data of the battery under the influence of the external environment data or external input data. Optionally, the battery excitation data includes the real-time SOC of the power battery during the entire charging process, the total current of the power battery pack, the real-time input ampere-hour of this charging, and the real-time ambient temperature, etc. The battery response data includes the total voltage of the power battery pack during the entire charging process, the voltage of each cell of the power battery pack, the average cell voltage of the power battery pack, the highest cell voltage of the power battery pack, the lowest cell voltage of the power battery pack, and the time series data of the temperature of each battery probe, etc.

[0082] Step 203: Input the battery excitation data and the battery response data into a first preset model to obtain first state characteristic data.

[0083] The battery excitation data and the battery response data are input into a first preset model to obtain first state characteristic data of a single vehicle single charging process.

[0084] In the above embodiment, considering the characteristics of the excitation-response system of the power battery system, the first battery status data is divided into system excitation data and system response data, and then input into the first preset model, which can better simulate the actual function of the power battery system, thereby improving the accuracy of the battery abnormality warning.

[0085] Optional, such as Figure 3 As shown, the first preset model includes a system encoder and a system decoder, and the battery excitation data and the battery response data are input into the first preset model to obtain the first state characteristic data, including:

[0086] Step 301: Input the battery excitation data and the battery response data into the system encoder to obtain the first encoded data.

[0087] There are many options for the specific network structures of the system encoder and system decoder, as long as they meet the following conditions: they can take time series data as input and can learn the correlation between multiple feature dimensions of the time series. Optional models include recurrent neural networks (RNN), long short-term memory neural networks (LSTM), gated recurrent unit networks (GRU), and full-state attention mechanism networks (Transformer encoders and Transformer decoders).

[0088] The battery excitation data and the battery response data are taken together as input data of the system encoder. After the system encoder encodes the input data, first encoded data is obtained.

[0089] Step 302: input the first coded data and the battery excitation data into the system decoder to obtain the first response data.

[0090] The first coded data and the battery excitation data are used as input data of a system decoder, and the system decoder processes the input data to obtain first response data reconstructed by a first preset model.

[0091] Step 303: determining first error data according to the first response data and the battery response data.

[0092] The difference between the first response data and the original battery response data is calculated to obtain first error data.

[0093] Step 304, obtaining first state characteristic data according to the first coded data and the first error data.

[0094] After the first coded data and the first error data are concatenated in dimension, the first state characteristic data corresponding to the single-charge timing data of a single vehicle is obtained.

[0095] In the above embodiment, the battery excitation data and the battery response data are used by the system encoder to extract the state characteristics of the battery system to obtain the first encoded data. The system decoder obtains the first response data based on the state characteristics extracted by the system encoder by simulating the function of the actual battery system, thereby obtaining the first state characteristic data, thereby realizing the state characteristic extraction of the battery single charging process.

[0096] For example, Figure 4 As shown, the training process of the first preset model includes:

[0097] Step 401 , obtaining sample battery status data of a sample vehicle, and determining second battery status data corresponding to each charging process of the sample vehicle according to the sample battery status data.

[0098] The sample vehicle may be one or more vehicles, and the second battery status data corresponding to each charging process of each sample vehicle is determined according to the sample battery status data. Multiple charging processes correspond to multiple second battery status data.

[0099] Step 402, determining sample battery excitation data and sample battery response data from the second battery status data.

[0100] Step 403: input the sample battery excitation data and the sample battery response data into an initial system encoder to obtain first sample encoded data.

[0101] Step 404: input the first sample coded data and the sample battery excitation data into an initial system decoder to obtain first sample decoded data.

[0102] Step 405 , calculating first sample error data according to the first sample decoded data and the sample battery response data.

[0103] Wherein, based on the first sample decoded data and the original sample battery response data, the reconstruction loss of the first preset model is calculated by a mean square error value such as an MSE value to obtain the first sample error data.

[0104] Step 406: Train the initial system encoder and the initial system decoder according to the first sample error data to obtain a system encoder and a system decoder.

[0105] Then, the parameters of the initial system encoder and the initial system decoder are trained by the back propagation algorithm (BP algorithm) until the reconstruction loss of the first preset model reaches the set value, and the training is completed to obtain the trained system encoder and system decoder, that is, the first preset model. Optionally, the network structure diagram of the first preset model can be as follows Figure 5 shown.

[0106] Exemplarily, in one embodiment, a gated recurrent unit network (GRU) is used as the specific network structure of the system encoder and the system decoder of the first preset model. The gated recurrent unit network is a recurrent neural network that receives a time series As input, that is, the time series of battery status data of this application, the data of each time step and the hidden state of the previous step inside the network Input the gated recurrent unit together to get the hidden state of the current step , then and the data for the next time step Input the gated recurrent unit again to get the hidden state of the next step , and so on.

[0107] like Figure 6 As shown in Figure 1, the gated recurrent unit itself is mainly composed of a reset gate, an update gate, and a candidate hidden state layer. The reset gate and the update gate are both composed of a fully connected layer connected to a sigmoid activation function, and the candidate hidden state layer is composed of a fully connected layer and a tanh activation function. The data at each time step and the hidden state of the previous step After splicing, input the reset gate and update gate to get the output of the reset gate The output of the update gate . Reset the output of the gate The hidden state of the previous step After element-wise multiplication, After splicing, input the candidate hidden state layer to obtain the candidate hidden state . Subsequently, the gate output is updated and Perform element-wise multiplication and and Perform element-wise multiplication and add the two results to get the hidden state of the current step The above calculation process is organized into the following formula:

[0108]

[0109] in and To reset the fully connected layer weight parameters of the gate, To reset the bias parameters of the fully connected layer of the gate, and is the weight parameter of the fully connected layer of the update gate, To update the bias parameters of the fully connected layer of the gate, and is the weight parameter of the candidate hidden state layer, is the bias parameter of the candidate hidden state layer, Represents the element-wise multiplication of two matrices.

[0110] In one embodiment, the second preset model includes a timing encoder and a timing decoder, and the steps of inputting a plurality of first state characteristic data into the second preset model to obtain the second state characteristic data of the vehicle are as follows: Figure 7 As shown, including:

[0111] Step 701, splicing multiple first state feature data according to time to obtain first time series data.

[0112] According to the above step 304, the first state characteristic data corresponding to each charging process of a single vehicle in a target charging cycle, such as a whole life cycle, is obtained, and the first state characteristic data corresponding to all charging processes of the same vehicle are spliced ​​into first time series data according to time.

[0113] Step 702, input the first time series data into a time series encoder to obtain second encoded data.

[0114] The first time series data is used as input data of the time series encoder for encoding to obtain second encoded data. The specific network structure of the time series encoder and the time series decoder can be the same as the network structure of the system encoder and the system decoder introduced above, or the gated recurrent unit network introduced above can be used.

[0115] Step 703: input the second encoded data into a timing decoder to obtain second decoded data.

[0116] The second encoded data is used as input data of the time series decoder and reconstructed to obtain second decoded data.

[0117] Step 704, determining second error data according to the second decoded data and the first timing data.

[0118] An error between the second decoded data and the original first time series data is calculated to obtain second error data.

[0119] Step 705: Use the second coded data and the second error data as second state feature data.

[0120] In the above embodiment, the first state characteristic data extracted by the first preset model is spliced ​​in time to obtain data of the power battery in the entire life cycle, and then the second state characteristic data is processed by the second preset model to achieve long-term feature extraction and change analysis of the power battery, so as to better obtain the signs of abnormal battery status.

[0121] Optionally, the training process of the second preset model is as follows Figure 8 As shown, including:

[0122] Step 801, obtaining sample state characteristic data according to first sample coding data and first sample error data.

[0123] According to the first sample encoding data and the first sample error data obtained in the first preset model training process, sample state feature data is obtained. One sample state feature data corresponds to the state feature of a single sample vehicle in a single charging process.

[0124] Step 802, splicing the sample state feature data according to time to obtain second time series data.

[0125] Multiple sample state characteristic data of the charging process of a single sample vehicle throughout its life cycle are spliced ​​according to actual conditions to obtain second time series data, which is also the time series characteristic data of the charging process of a single sample vehicle throughout its life cycle.

[0126] Step 803: input the second time series data into the initial time series encoder to obtain second sample encoded data.

[0127] Step 804: input the second sample encoded data into the initial timing decoder to obtain second sample decoded data.

[0128] Step 805 , determining second sample error data according to the second sample decoded data and the second timing data.

[0129] The second sample decoded data and the second time series data are used to calculate the reconstruction loss of the second preset model through a mean square error value such as an MSE value to obtain second sample error data.

[0130] Step 806: Train the initial temporal encoder and the initial temporal decoder according to the second sample error data to obtain a temporal encoder and a temporal decoder.

[0131] Then, the parameters of the initial temporal encoder and the initial temporal decoder are trained by a back propagation algorithm such as the BP algorithm until the reconstruction loss of the second preset model reaches a set value, and the training is terminated to obtain the temporal encoder and the temporal decoder after the training is completed, that is, the second preset model. Optionally, the network structure diagram of the second preset model can be as follows Fig. 9 shown.

[0132] In one embodiment, the battery abnormality warning result of the vehicle is determined based on the second state characteristic data, including the following two situations.

[0133] In the first case, the battery status data of a vehicle with an abnormal battery in a thermal runaway case, the method for determining the battery abnormality warning result of the vehicle includes: inputting the second status characteristic data into a third preset model, and determining the battery abnormality warning result according to the output result of the third preset model.

[0134] Among them, the third preset model is trained based on the second state characteristic data of the sample vehicle. The second state characteristic data of all vehicles are labeled with whether the battery has thermal runaway, and then the third preset model is constructed. The third preset model can be a machine learning model, such as traditional machine learning models such as random forest, xgboost, etc., or a fully connected neural network model. The second state characteristic data of all vehicles including the second coding data and the second error data are trained with the thermal runaway label data to obtain the third preset model. During the detection process, the second state characteristic data is input into the third preset model, and the thermal runaway warning score of each vehicle is obtained according to the output result of the third preset model, thereby determining the abnormal battery warning result of each vehicle.

[0135] In the second case, the battery status data of a vehicle with abnormal battery without a thermal runaway case, the method for determining the battery abnormality warning result of the vehicle includes: determining the battery abnormality warning result according to the second error data in the second state characteristic data and the battery abnormality warning threshold.

[0136] The battery abnormality warning threshold is calculated based on the mean and standard deviation of the second error data of the sample vehicles. With standard deviation , calculate the battery abnormality warning threshold During the detection process, when the second error data exceeds the battery abnormality warning threshold, it is determined that the battery is abnormal and a thermal runaway warning is issued.

[0137] In the embodiments of the present application, Fig.10 FIG. 1 is a flow chart of a battery abnormality warning method provided in an embodiment of the present application, including:

[0138] Step 1001, obtaining battery status data of a vehicle within a target charging cycle.

[0139] Step 1002: Determine first battery status data corresponding to each charging process of the vehicle according to the battery status data.

[0140] Step 1003, determining battery excitation data and battery response data from the first battery status data.

[0141] Step 1004: input the battery excitation data and the battery response data into a first preset model to obtain first state characteristic data.

[0142] Step 1005: splice multiple first state feature data according to time to obtain first time series data.

[0143] Step 1006, input the first time series data into a time series encoder to obtain second encoded data.

[0144] Step 1007: input the second encoded data into a timing decoder to obtain second decoded data.

[0145] Step 1008, determining second error data according to the second decoded data and the first timing data.

[0146] Step 1009: Use the second coded data and the second error data as second state feature data.

[0147] Step 1010, determining a battery abnormality warning result of the vehicle according to the second state characteristic data.

[0148] In the above embodiment, the battery features are extracted and analyzed through two layers of models, namely the first preset model and the second preset model. By analyzing the features of the short-cycle data and the long-cycle data of the battery, it is possible to better warn of abnormalities such as battery thermal runaway that are difficult to warn through short-term data, but which may have certain signs in long-term data. The accuracy of battery abnormality warning is higher.

[0149] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0150] Based on the same inventive concept, the embodiment of the present application also provides a battery abnormality warning device for implementing the battery abnormality warning method involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in one or more battery abnormality warning device embodiments provided below can refer to the limitations of the battery abnormality warning method above, and will not be repeated here.

[0151] In an exemplary embodiment, Fig.11 As shown, a battery abnormality warning device 1100 is provided, comprising: an acquisition module, a first input module, a second input module and a determination module, wherein:

[0152] An acquisition module, used to acquire battery status data of a vehicle within a target charging cycle, where the target charging cycle includes multiple charging processes;

[0153] A first input module, used to input the battery state data into a first preset model to obtain a plurality of first state characteristic data; the first state characteristic data is used to characterize the state characteristics corresponding to each charging process of the vehicle within a target charging cycle;

[0154] A second input module is used to input a plurality of first state characteristic data into a second preset model to obtain second state characteristic data; the second state characteristic data is used to characterize state characteristics corresponding to multiple charging processes of the vehicle within a target charging cycle;

[0155] The determination module is used to determine the battery abnormality warning result of the vehicle according to the second state characteristic data.

[0156] In one embodiment, the first input module is specifically used to determine the first battery status data corresponding to each charging process of the vehicle based on the battery status data; determine the battery excitation data and the battery response data from the first battery status data, the battery excitation data is used to characterize the external environment data or external input data of the battery during the charging process, and the battery response data is used to characterize the response data of the battery under the influence of the external environment data or external input data; input the battery excitation data and the battery response data into the first preset model to obtain the first state characteristic data.

[0157] In one embodiment, the first preset model includes a system encoder and a system decoder, and the first input module is specifically used to input battery excitation data and battery response data into the system encoder to obtain first encoded data; input the first encoded data and the battery excitation data into the system decoder to obtain first response data; determine first error data based on the first response data and the battery response data; and obtain first state characteristic data based on the first encoded data and the first error data.

[0158] In one embodiment, the second preset model includes a timing encoder and a timing decoder, and the second input module is specifically used to splice multiple first state feature data according to time to obtain first time series data; input the first time series data into the timing encoder to obtain second encoded data; input the second encoded data into the timing decoder to obtain second decoded data; determine second error data based on the second decoded data and the first time series data; and use the second encoded data and the second error data as second state feature data.

[0159] In one embodiment, the determination module is specifically used to input the second state characteristic data into a third preset model, and determine the battery abnormality warning result according to the output result of the third preset model. The third preset model is trained based on the second state characteristic data of the sample vehicle, and the sample vehicle includes a vehicle with battery abnormality.

[0160] In one of the embodiments, the determination module is also used to determine the battery abnormality warning result based on the second error data in the second state characteristic data and the battery abnormality warning threshold; wherein the battery abnormality warning threshold is calculated based on the mean and standard deviation of the second error data of the sample vehicles, and the sample vehicles do not include vehicles with battery abnormalities.

[0161] In one embodiment, the device also includes a first training module, which is used to obtain sample battery status data of a sample vehicle, determine second battery status data corresponding to each charging process of the sample vehicle based on the sample battery status data; determine sample battery excitation data and sample battery response data from the second battery status data; input the sample battery excitation data and the sample battery response data into an initial system encoder to obtain first sample encoded data; input the first sample encoded data and the sample battery excitation data into an initial system decoder to obtain first sample decoded data; calculate first sample error data based on the first sample decoded data and the sample battery response data; and train the initial system encoder and the initial system decoder based on the first sample error data to obtain a system encoder and a system decoder.

[0162] In one embodiment, the device also includes a second training module, which is used to obtain sample state characteristic data based on the first sample encoding data and the first sample error data; splicing the sample state characteristic data according to time to obtain second time series data; inputting the second time series data into the initial time series encoder to obtain second sample encoding data; inputting the second sample encoding data into the initial time series decoder to obtain second sample decoded data; determining the second sample error data based on the second sample decoded data and the second time series data; and training the initial time series encoder and the initial time series decoder based on the second sample error data to obtain the time series encoder and the time series decoder.

[0163] Each module in the above battery abnormality warning device can be implemented in whole or in part by software, hardware and their combination. Each module can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the corresponding operations of each module.

[0164] In an exemplary embodiment, a computer device is provided. The computer device may be a terminal, and its internal structure diagram may be as shown in FIG. Fig.12As shown. The computer device includes a processor, a memory, an input / output interface, a communication interface, a display unit and an input device. The processor, the memory and the input / output interface are connected through a system bus, and the communication interface, the display unit and the input device are connected to the system bus through the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be implemented through WIFI, a mobile cellular network, near field communication (NFC) or other technologies. When the computer program is executed by the processor, a battery abnormality warning method is implemented. The display unit of the computer device is used to form a visually visible picture, which can be a display screen, a projection device or a virtual reality imaging device. The display screen can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covering the display screen, or a button, trackball or touchpad set on the computer device shell, or an external keyboard, touchpad or mouse.

[0165] Those skilled in the art will understand that Fig.12 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0166] In an exemplary embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the following steps when executing the computer program: obtaining battery status data of a vehicle within a target charging cycle, wherein the target charging cycle includes multiple charging processes; inputting the battery status data into a first preset model to obtain multiple first status characteristic data; the first status characteristic data is used to characterize status characteristics corresponding to each charging process of the vehicle within the target charging cycle; inputting the multiple first status characteristic data into a second preset model to obtain second status characteristic data; the second status characteristic data is used to characterize status characteristics corresponding to multiple charging processes of the vehicle within the target charging cycle; and determining a battery abnormality warning result of the vehicle based on the second status characteristic data.

[0167] In one embodiment, the processor further implements the following steps when executing the computer program: determining first battery status data corresponding to each charging process of the vehicle based on the battery status data; determining battery excitation data and battery response data from the first battery status data, the battery excitation data being used to characterize external environmental data or external input data of the battery during the charging process, and the battery response data being used to characterize the response data of the battery under the influence of external environmental data or external input data; and inputting the battery excitation data and the battery response data into a first preset model to obtain first state characteristic data.

[0168] In one embodiment, the first preset model includes a system encoder and a system decoder, and the processor also implements the following steps when executing the computer program: inputting the battery excitation data and the battery response data into the system encoder to obtain first coded data; inputting the first coded data and the battery excitation data into the system decoder to obtain first response data; determining the first error data based on the first response data and the battery response data; and obtaining the first state characteristic data based on the first coded data and the first error data.

[0169] In one embodiment, the second preset model includes a timing encoder and a timing decoder, and the processor also implements the following steps when executing the computer program: splicing multiple first state feature data according to time to obtain first timing data; inputting the first timing data into the timing encoder to obtain second encoded data; inputting the second encoded data into the timing decoder to obtain second decoded data; determining second error data based on the second decoded data and the first timing data; and using the second encoded data and the second error data as second state feature data.

[0170] In one embodiment, when the processor executes the computer program, the following steps are also implemented: inputting the second state characteristic data into a third preset model, and determining the battery abnormality warning result according to the output result of the third preset model, wherein the third preset model is trained based on the second state characteristic data of the sample vehicle, and the sample vehicle includes a vehicle with a battery abnormality.

[0171] In one embodiment, when the processor executes the computer program, the following steps are also implemented: determining the battery abnormality warning result based on the second error data in the second state characteristic data and the battery abnormality warning threshold; wherein the battery abnormality warning threshold is calculated based on the mean and standard deviation of the second error data of the sample vehicles, and the sample vehicles do not include vehicles with battery abnormalities.

[0172] In one embodiment, when the processor executes the computer program, the following steps are also implemented: obtaining sample battery status data of the sample vehicle, and determining the second battery status data corresponding to each charging process of the sample vehicle based on the sample battery status data; determining sample battery excitation data and sample battery response data from the second battery status data; inputting the sample battery excitation data and the sample battery response data into an initial system encoder to obtain first sample encoded data; inputting the first sample encoded data and the sample battery excitation data into an initial system decoder to obtain first sample decoded data; calculating first sample error data based on the first sample decoded data and the sample battery response data; and training the initial system encoder and the initial system decoder based on the first sample error data to obtain a system encoder and a system decoder.

[0173] In one embodiment, when the processor executes the computer program, the following steps are also implemented: sample state characteristic data is obtained based on the first sample encoding data and the first sample error data; the sample state characteristic data is spliced ​​according to time to obtain second time series data; the second time series data is input into the initial time series encoder to obtain second sample encoding data; the second sample encoding data is input into the initial time series decoder to obtain second sample decoded data; the second sample error data is determined based on the second sample decoded data and the second time series data; the initial time series encoder and the initial time series decoder are trained based on the second sample error data to obtain the time series encoder and the time series decoder.

[0174] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the battery abnormality warning method described in any of the above embodiments is implemented.

[0175] In one embodiment, a computer program product is provided, including a computer program, which, when executed by a processor, implements the battery abnormality warning method described in any one of the above embodiments.

[0176] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0177] A person of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment method can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present 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. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic unit, a data processing logic unit based on quantum computing, an artificial intelligence (AI) processor, etc., but are not limited to this.

[0178] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, 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.

[0179] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be construed as limiting the scope of the present application. It should be noted that, for a person of ordinary skill in the art, several modifications and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A battery abnormality warning method, characterized in that: The method comprises: Acquiring battery status data of a vehicle within a target charging cycle, wherein the target charging cycle includes multiple charging processes; Inputting the battery status data into a first preset model to obtain a plurality of first status characteristic data; the first status characteristic data is used to characterize the status characteristics corresponding to each charging process of the vehicle within the target charging cycle; Inputting the plurality of first state characteristic data into a second preset model to obtain second state characteristic data; the second state characteristic data is used to characterize state characteristics corresponding to multiple charging processes of the vehicle within the target charging cycle; Determining a battery abnormality warning result of the vehicle according to the second state characteristic data; The second preset model includes a timing encoder and a timing decoder, and the inputting of the plurality of first state characteristic data into the second preset model to obtain the second state characteristic data of the vehicle includes: The plurality of first state feature data are spliced ​​together according to time to obtain first time series data; Inputting the first time series data into the time series encoder to obtain second encoded data; Inputting the second encoded data into the timing decoder to obtain second decoded data; determining second error data according to an error between the second decoded data and the first timing data; using the second coded data and the second error data as the second state characteristic data; The determining, according to the second state characteristic data, a battery abnormality warning result of the vehicle includes: Inputting the second state characteristic data into a third preset model, and determining the battery abnormality warning result according to an output result of the third preset model, wherein the third preset model is trained based on the second state characteristic data of a sample vehicle, and the sample vehicle includes a vehicle with a battery abnormality; or, The battery abnormality warning result is determined based on the second error data in the second state characteristic data and the battery abnormality warning threshold; wherein the battery abnormality warning threshold is calculated based on the mean and standard deviation of the second error data of sample vehicles, and the sample vehicles do not include vehicles with battery abnormalities.

2. The method according to claim 1, characterized in that The step of inputting the battery status data into a first preset model to obtain a plurality of first status characteristic data includes: Determining first battery status data corresponding to each charging process of the vehicle according to the battery status data; Determine battery excitation data and battery response data from the first battery state data, wherein the battery excitation data is used to characterize external environmental data or external input data of the battery during charging, and the battery response data is used to characterize response data of the battery under the influence of the external environmental data or the external input data; The battery excitation data and the battery response data are input into the first preset model to obtain the first state characteristic data.

3. The method according to claim 2, characterized in that The first preset model includes a system encoder and a system decoder, and the step of inputting the battery excitation data and the battery response data into the first preset model to obtain the first state characteristic data includes: Inputting the battery excitation data and the battery response data into the system encoder to obtain first encoded data; Inputting the first coded data and the battery excitation data into the system decoder to obtain first response data; determining first error data according to a difference between the first response data and the battery response data; The first state characteristic data is obtained according to the first coded data and the first error data.

4. The method according to claim 3, characterized in that: The training process of the first preset model includes: Acquire sample battery status data of a sample vehicle, and determine second battery status data corresponding to each charging process of the sample vehicle according to the sample battery status data; determining sample battery stimulus data and sample battery response data from the second battery state data; Inputting the sample battery excitation data and the sample battery response data into an initial system encoder to obtain first sample encoded data; Inputting the first sample coded data and the sample battery excitation data into an initial system decoder to obtain first sample decoded data; Calculating first sample error data according to the first sample decoded data and the sample battery response data; The initial system encoder and the initial system decoder are trained according to the first sample error data to obtain the system encoder and the system decoder.

5. The method according to claim 4, characterized in that The training process of the second preset model includes: Obtaining sample state characteristic data according to the first sample encoding data and the first sample error data; The sample state characteristic data is spliced ​​according to time to obtain second time series data; Inputting the second time series data into an initial time series encoder to obtain second sample encoded data; Inputting the second sample encoded data into an initial timing decoder to obtain second sample decoded data; Determine second sample error data according to the second sample decoded data and the second timing data; The initial temporal encoder and the initial temporal decoder are trained according to the second sample error data to obtain a temporal encoder and a temporal decoder.

6. A battery abnormality warning device, characterized in that: The device comprises: An acquisition module, used to acquire battery status data of a vehicle within a target charging cycle, wherein the target charging cycle includes multiple charging processes; A first input module, used for inputting the battery status data into a first preset model to obtain a plurality of first status characteristic data; the first status characteristic data is used for characterizing the status characteristics corresponding to each charging process of the vehicle within the target charging cycle; A second input module, used for inputting the plurality of first state characteristic data into a second preset model to obtain second state characteristic data; the second state characteristic data is used for characterizing state characteristics corresponding to the plurality of charging processes of the vehicle within the target charging cycle; A determination module, configured to determine a battery abnormality warning result of the vehicle according to the second state characteristic data; The second preset model includes a timing encoder and a timing decoder, and the second input module is specifically used to splice the plurality of first state feature data according to time to obtain first time series data; input the first time series data into the timing encoder to obtain second encoded data; input the second encoded data into the timing decoder to obtain second decoded data; determine second error data according to an error between the second decoded data and the first time series data; and use the second encoded data and the second error data as the second state feature data; The determination module is specifically used to input the second state characteristic data into a third preset model, and determine the battery abnormality warning result according to the output result of the third preset model, wherein the third preset model is trained based on the second state characteristic data of a sample vehicle, and the sample vehicle includes a vehicle with a battery abnormality; or, is specifically used to determine the battery abnormality warning result based on the second error data in the second state characteristic data and a battery abnormality warning threshold; wherein the battery abnormality warning threshold is calculated based on the mean and standard deviation of the second error data of the sample vehicle, and the sample vehicle does not include a vehicle with a battery abnormality.

7. The device according to claim 6, characterized in that The first input module is specifically used to determine the first battery status data corresponding to each charging process of the vehicle according to the battery status data; Determine battery excitation data and battery response data from the first battery state data, wherein the battery excitation data is used to characterize external environmental data or external input data of the battery during charging, and the battery response data is used to characterize response data of the battery under the influence of the external environmental data or the external input data; The battery excitation data and the battery response data are input into the first preset model to obtain the first state characteristic data.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 5 are implemented.

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

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

Citation Information

Patent Citations

  • Power battery consistency safety state evaluation method

    CN114879049A