A reliability evaluation method and system for a battery thermal management system

Abstract

The present invention provides a method and system for evaluating the reliability of a battery thermal management system. By using a heat dissipation performance prediction model and a prediction result verification model, the temperature distribution under future battery discharge and charging conditions can be predicted, potential future battery high temperatures can be warned, and the probability of battery thermal runaway can be reduced. By adopting a parallel encoder structure to process the discharge data of different battery stages respectively, the accuracy of predicting future temperature time series data can be improved. By cross-verifying the preliminary performance prediction results output by the heat dissipation performance prediction model and the prediction result verification model, the robustness of this evaluation method is greatly improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of battery thermal management, and particularly relates to a method and a system for evaluating the reliability of a battery thermal management system. Background Art

[0002] The statements in this part only provide background technical information related to the present invention and do not necessarily constitute prior art.

[0003] In the practical application of electric vehicles, the battery thermal management system aimed at ensuring the safe operation of the battery is crucial. The temperature state change of lithium-ion batteries has a great impact on their performance release and operation safety. A reliable battery thermal management system can ensure that the battery operates under safe conditions and reduce the rapid decay of the battery life caused by thermal abuse. Therefore, it is crucial to develop an effective method for evaluating the reliability of the battery thermal management system.

[0004] At present, the evaluation methods for the reliability of battery thermal management systems mainly include the following categories:

[0005] 1) By comparing the real-time monitoring values of the working medium operation temperature and flow rate of the battery thermal management system with the preset operation experience values, analyzing the real-time operation state of the battery thermal management system to evaluate the reliability of the battery thermal management system. This evaluation method based on operation experience values has low accuracy. Different battery thermal management systems have different operation experience values and need to be retested, so its portability is also poor.

[0006] 2) Using the battery temperature threshold feedback mechanism, by comparing the real-time temperature monitoring value of the battery obtained by the temperature sensor with the preset battery safe operation temperature threshold to evaluate the reliability of the battery thermal management system. When this method triggers an over-temperature warning, the battery operating temperature is relatively high. If the battery cannot be cooled in a timely and effective manner, it is likely to lead to battery thermal runaway, and the cost is relatively high.

[0007] Based on the defects of the above methods, there is an urgent need for a method for evaluating the reliability of a battery thermal management system with low cost, high accuracy and good portability to evaluate the working state of the battery thermal management system, so as to ensure that the battery works in a suitable temperature range for a long time. Summary of the Invention

[0008] To overcome the deficiencies of the above-mentioned prior art, the present invention provides a method and a system for evaluating the reliability of a battery thermal management system. By using a heat dissipation performance prediction model and a prediction result verification model, the temperature distribution under future battery discharge and charge conditions can be predicted, and potential future battery high temperatures can be warned, reducing the probability of battery thermal runaway.

[0009] To achieve the above object, the first aspect of the present invention provides a method for evaluating the reliability of a battery thermal management system, including:

[0010] Obtain the discharge timing data and real-time charging temperature timing data at different stages of the temperature measurement points of the battery thermal management system. Among them, the discharge timing data at different stages includes: real-time discharge temperature timing data, SOC timing data, initial cycle discharge temperature timing data, SOC timing data; and the temperature difference timing data and SOC difference timing data of the adjacent cycle discharge conditions before the real-time operating conditions.

[0011] Input the discharge timing data and the real-time charging temperature timing data into the trained heat dissipation performance prediction model and prediction result verification model respectively to obtain the discharge temperature prediction result and the charging temperature prediction result; evaluate the reliability of the battery thermal management system based on the prediction results.

[0012] Among them, the heat dissipation performance prediction model includes parallel encoders. The parallel encoders are used to extract features from the discharge timing data at different stages respectively, and the extracted features are fused and then decoded to obtain the discharge temperature prediction result.

[0013] The second aspect of the present invention provides a battery thermal management system reliability evaluation system, including:

[0014] An acquisition module, configured to obtain the discharge timing data and the real-time charging temperature timing data at different stages of the temperature measurement points of the battery thermal management system. Among them, the discharge timing data at different stages includes: real-time discharge temperature timing data, SOC timing data, initial cycle discharge temperature timing data, SOC timing data; and the temperature difference timing data and SOC difference timing data of the adjacent cycle discharge conditions before the real-time operating conditions.

[0015] An evaluation module, configured to input the discharge timing data and the real-time charging temperature timing data into the trained heat dissipation performance prediction model and prediction result verification model respectively to obtain the discharge temperature prediction result and the charging temperature prediction result; evaluate the reliability of the battery thermal management system based on the prediction results.

[0016] Among them, the heat dissipation performance prediction model includes parallel encoders. The parallel encoders are used to extract features from the discharge timing data at different stages respectively, and the extracted features are fused and then decoded to obtain the discharge temperature prediction result.

[0017] The third aspect of the present invention provides a computer device, including: a processor, a memory and a bus. The memory stores machine-readable instructions executable by the processor. When the computer device runs, the processor communicates with the memory through the bus. When the machine-readable instructions are executed by the processor, a battery thermal management system reliability evaluation method is executed.

[0018] The fourth aspect of the present invention provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is run by a processor, it executes a method for evaluating the reliability of a battery thermal management system.

[0019] The above one or more technical solutions have the following beneficial effects:

[0020] In the present invention, a heat dissipation performance prediction model and a prediction result verification model are used to predict the temperature distribution under future battery discharge and charging conditions, which can give early warnings of potential future battery high temperatures and reduce the probability of battery thermal runaway; moreover, the method of the present invention does not need to measure operation experience values for different battery thermal management systems and has the advantage of good portability.

[0021] In the present invention, a parallel encoder structure is adopted to process real-time discharge temperature time series data and SOC time series data; process the initial cycle discharge temperature time series data and SOC time series data, and the battery temperature and SOC distribution characteristics when the battery thermal management system is in good condition can be obtained; and by processing the temperature difference time series data and SOC difference time series data of the adjacent cycle discharge conditions before the real-time operating condition, the characteristics representing the performance degradation trend of the battery thermal management system can be obtained. By fusing the characteristics and decoding for prediction, the accuracy of predicting future temperature time series data can be improved.

[0022] The advantages of the additional aspects of the present invention will be partially given in the following description, partially become obvious from the following description, or be understood through the practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] The specification drawings constituting a part of the present invention are used to provide a further understanding of the present invention. The schematic embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation to the present invention.

[0024] Figure 1 It is a flowchart of the method for evaluating the reliability of the battery thermal management system in Embodiment 1 of the present invention;

[0025] Figure 2 It is a program block diagram of the method for evaluating the reliability of the battery thermal management system based on the improved Transformer in Embodiment 1 of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0026] It should be noted that the following detailed description is exemplary and is intended to provide further illustration of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by those of ordinary skill in the technical field to which the present invention belongs.

[0027] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present invention.

[0028] In the case of no conflict, the embodiments in the present invention and the features in the embodiments can be combined with each other.

[0029] Embodiment 1

[0030] This embodiment discloses a method for evaluating the reliability of a battery thermal management system, including:

[0031] Obtain the discharge timing data and real-time charging temperature timing data of the temperature measurement points of the battery thermal management system at different stages. Among them, the discharge timing data at different stages includes: real-time discharge temperature timing data, SOC timing data, initial cycle discharge temperature timing data, SOC timing data; and the temperature difference timing data and SOC difference timing data of the adjacent cycle discharge conditions before the real-time operating condition.

[0032] Input the discharge timing data and real-time charging temperature timing data into the trained heat dissipation performance prediction model and prediction result verification model respectively to obtain the discharge temperature prediction result and the charging temperature prediction result; evaluate the reliability of the battery thermal management system based on the prediction results.

[0033] Among them, the heat dissipation performance prediction model includes parallel encoders. The parallel encoders are used to extract features from the discharge timing data at different stages respectively, and the extracted features are fused and then decoded to obtain the discharge temperature prediction result.

[0034] This embodiment uses the heat dissipation performance prediction model and the prediction result verification model to predict the temperature distribution under future battery discharge and charging conditions, which can warn of potential future battery high temperatures and reduce the probability of battery thermal runaway; moreover, the method of the present invention does not need to measure the operating experience values for different battery thermal management systems and has the advantage of good portability.

[0035] This embodiment adopts a parallel encoder structure. By processing the real-time discharge temperature timing data and SOC timing data; processing the initial cycle discharge temperature timing data and SOC timing data, the battery temperature and SOC distribution characteristics when the battery thermal management system is in good condition can be obtained; and by processing the temperature difference timing data and SOC difference timing data of the adjacent cycle discharge conditions before the real-time operating condition, the characteristics representing the performance decay trend of the battery thermal management system can be obtained. By fusing the features and decoding for prediction, the accuracy of predicting future temperature time series data can be improved.

[0036] Combine Figure 1 and Figure 2A detailed description is given of the reliability evaluation of a battery thermal management system proposed in this embodiment:

[0037] S1: Select the liquid cooling working medium of the battery thermal management system and the temperature measurement points on the battery surface, and collect the temperature data of each measurement point, the state of charge SOC of the battery, and the corresponding measurement time.

[0038] Among them, the temperature measurement points are selected as the three temperature points of the inlet Tin, the middle position T1, and the outlet Tout of the cold plate flow channel of the battery thermal management system, and the three temperature points of the battery surface temperature Tm-in, Tm-1, and Tm-out corresponding to the above temperature point positions.

[0039] The measured data includes multiple charge and discharge cycle data. One charge cycle is the process from the start of battery charging to the maximum value of SOC, and one discharge cycle is the process from the start of battery discharge to the minimum value of SOC.

[0040] S2: Divide the collected data according to different charge and discharge conditions and perform data processing to obtain training samples, which specifically include the following steps:

[0041] S21: Divide the data collected in S1 into a discharge condition database B discharge and a charge condition heat dissipation performance database B charge . These two databases respectively contain multiple complete discharge cycle data and charge cycle data.

[0042] S22: Extract the discharge condition data of the first cycle batch from the discharge condition database B discharge as the input data B base1 for the heat dissipation performance prediction model.

[0043] The input data B base1 for the heat dissipation performance prediction model contains all the temperature-time series data corresponding to each temperature point from the start of discharge to the end of discharge in the first discharge cycle and the SOC-time series data of the battery, and is the fixed input of the heat dissipation performance prediction model.

[0044] S23: Extract the discharge condition data after the second cycle batch from the discharge condition database B discharge , and apply a sliding window to process the temperature sequence and SOC sequence therein. The length of the sliding window is N, and the step size of each movement is 1. Use the temperature sequence data and SOC sequence data within the sliding window range as the input data B base2 for the heat dissipation performance prediction model, and use the temperature sequence with a length of L located after the sliding window as the corresponding label Y. Among them, N and L are positive integers, and N > L.

[0045] S24: Assume that the sliding window in S23 is located in Bdischarge Among the data of the k-th cyclic batch, where k is a positive integer and k is incremented by 1 when the sliding window enters the next cyclic batch. Extract B discharge The discharge condition data of the first n cyclic batches in the k-th time in B

[0046] Taking the SOC difference as the benchmark, for the temperature data at the same temperature measurement point among the discharge condition data of n cyclic batches, subtract the temperature data of adjacent cycles batch by batch to obtain the temperature difference sequence between cyclic batches, and number them 1 to n in sequence from front to back according to the above subtraction batches to obtain the numbering order of the temperature difference sequence. Synthesize the temperature difference sequence in ascending order of batch numbers and in the time dimension, and use the synthesized data as the input B of the heat dissipation performance prediction model base3 , and this input only occurs in B in S23 base2 Changes only when entering the next cyclic batch

[0047] S25: Combine the above three inputs and labels to form a sample M for training the heat dissipation performance prediction model discharge ;

[0048] S26: Apply a sliding window to the temperature sequence and SOC sequence in the charging condition heat dissipation performance database B charge The length of the sliding window is N, and the step size of each movement is 1. Use the temperature and SOC sequences within the sliding window range as the input P for validating the model of the prediction result, and use the temperature sequence with a length of L located after the sliding window as the corresponding label. The input P and the corresponding label together form a sample M for training the heat dissipation performance prediction model charge .

[0049] S3: Based on the Transformer principle, establish a reliability evaluation framework for the battery thermal management system. This framework includes a heat dissipation performance prediction model based on an improved Transformer structure and a prediction result validation model based on the original Transformer structure

[0050] Specifically, it includes the following steps

[0051] S31: In this framework, the heat dissipation performance prediction model processes the sample M obtained from the discharge condition database B discharge , and the prediction result validation model processes the sample M obtained from the charging condition database B discharge . charge in the charge .

[0052] S32: Based on the traditional Transformer, make improvements to obtain the heat dissipation performance prediction model

[0053] Specifically, the structure of one encoder corresponding to one decoder is changed to the structure of three parallel encoders corresponding to one decoder. The three parallel encoders are the upper limit correction encoder, the main encoder, and the performance degradation encoder respectively.

[0054] S33: The upper limit correction encoder takes B base1 as the input and outputs the feature X1 containing the information about the battery temperature distribution when the battery thermal management system is in good condition.

[0055] The main encoder takes B base2 as the input and outputs the feature X2 containing the information about the latest temperature distribution and SOC.

[0056] The performance degradation encoder takes B base3 as the input and outputs the feature X3 containing the performance degradation trend of the battery thermal management system. The above features all exist in the form of a tensor space.

[0057] S34: Set weights for the three features obtained in S33 respectively:

[0058]

[0059]

[0060]

[0061] Multiply the weight coefficients by the corresponding weights and then sum them up to obtain the fused feature X.

[0062] Input X into the discharge condition decoder to predict the discharge temperature time series at future moments. q is the battery life reference value characterized by the number of cycles provided by the manufacturer, and l is the total number of charge cycles of the battery up to now;

[0063] S35: The prediction result verification model retains the traditional Transformer structure.

[0064] S4: Use the training samples obtained in S2 to train the two models in the architecture described in S3 respectively. Specifically, it includes the following steps:

[0065] S41: Use the training samples M discharge and M charge to train the heat dissipation performance prediction model and the prediction result verification model respectively. Each input B discharge in the training sample M base1 , B base2 , B base3 needs to be input from the corresponding encoders in the heat dissipation performance prediction model respectively.

[0066] S5: Process the real-time charge and discharge condition data and the discharge condition data of the last n cycle batches in the non-real-time data; after processing, input them into the two trained models to obtain the battery charging temperature prediction result and the battery discharging temperature prediction result.

[0067] Through the evaluation of the two results, the preliminary performance evaluation results under the charging condition and the preliminary performance evaluation results under the discharging condition are obtained respectively.

[0068] Compare the preliminary performance evaluation results under the charging condition with the preliminary performance evaluation results under the discharging condition to generate the reliability evaluation of the battery thermal management system. Specifically, it includes the following steps:

[0069] S51: Process the real-time discharge data and the discharge condition data of the latest two cycle batches in B discharge except the real-time discharge condition data according to the method of processing data in step S2, and then input them into the heat dissipation performance prediction model to predict the discharge temperature time series at future moments. If the predicted value of this temperature time series is higher than the measured temperature time series under the subsequent discharge condition, it is judged that the reliability is good; otherwise, it is judged that the reliability is reduced.

[0070] S52: Process the real-time charge data according to the method of processing data in step S2, and then input it into the prediction result verification model to predict the charge temperature time series at future moments. If the predicted value of this temperature time series is higher than the measured temperature time series under the subsequent charge condition, it is judged that the reliability is good; otherwise, it is judged that the reliability is reduced.

[0071] S53: When the judgments of the heat dissipation performance prediction model and the prediction result verification model are the same, it is output as the final reliability evaluation result; when the two judgments are different, due to the stronger stability of the input of the battery charging condition data, the output of the prediction result verification model is adopted as the reliability judgment.

[0072] The parallel architecture established in this embodiment greatly improves the robustness of this evaluation method by cross-verifying the preliminary performance prediction results output by the heat dissipation performance prediction model and the prediction result verification model.

[0073] The present invention selects the liquid cooling working medium of the battery thermal management system and the temperature measurement points on the battery surface, collects the temperature data of each temperature measurement point, the state of charge SOC of the battery, and the corresponding measurement time; divides the collected data according to different charge and discharge working conditions and performs data processing to obtain training samples; based on the Transformer principle, a reliability evaluation architecture for the battery thermal management system is established. This architecture includes a heat dissipation performance prediction model based on an improved Transformer structure and a prediction result verification model based on the original Transformer structure; uses the training samples obtained in S2 to train the two models in the architecture described in S3 respectively; processes the real-time charge and discharge working condition data and the last n cycle batch discharge working condition data in the non-real-time data. After processing, the data is input into the two trained models to obtain the battery charging temperature prediction result and the battery discharge temperature prediction result. Through the evaluation of the two results, the preliminary performance evaluation results under the charging working condition and the preliminary performance evaluation results under the discharge working condition are obtained respectively. By comparing the preliminary performance evaluation results under the charging working condition with the preliminary performance evaluation results under the discharge working condition, a reliability evaluation of the battery thermal management system is generated, which has the advantages of low cost, high accuracy, and good portability.

[0074] Embodiment 2

[0075] The purpose of this embodiment is to provide a reliability evaluation system for a battery thermal management system, including:

[0076] An acquisition module, configured to acquire the discharge timing data and the real-time charging temperature timing data at different stages of the temperature measurement points of the battery thermal management system. Among them, the discharge data at different stages includes: real-time discharge temperature timing data, SOC timing data, initial cycle discharge temperature timing data, SOC timing data; and the temperature difference timing data and SOC difference timing data of the cycle discharge working condition adjacent to the real-time operation working condition.

[0077] An evaluation module, configured to input the discharge timing data and the real-time charging temperature timing data into the trained heat dissipation performance prediction model and prediction result verification model respectively to obtain the discharge temperature prediction result and the charging temperature prediction result; and evaluate the reliability of the battery thermal management system based on the prediction results.

[0078] Among them, the heat dissipation performance prediction model includes parallel encoders, and the parallel encoders are used to extract features from the discharge timing data at different stages respectively, and the extracted features are fused and then decoded to obtain the discharge temperature prediction result.

[0079] Embodiment 3

[0080] The purpose of this embodiment is to provide a computing device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, the steps of the above method are implemented.

[0081] Example 4

[0082] The purpose of this embodiment is to provide a computer-readable storage medium.

[0083] A computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, it performs the steps of the above method.

[0084] The steps involved in the devices of the above Examples 2, 3 and 4 correspond to those of Method Example 1, and for the specific implementation manners, reference may be made to the relevant description part of Example 1. The term "computer-readable storage medium" should be understood to include a single medium or multiple media including one or more instruction sets; it should also be understood to include any medium that can store, encode or carry an instruction set for execution by a processor and enable the processor to execute any method in the present invention.

[0085] Those skilled in the art should understand that the above-mentioned modules or steps of the present invention can be implemented by a general-purpose computer device. Optionally, they can be implemented by program codes executable by a computing device, so that they can be stored in a storage device and executed by the computing device, or they can be separately fabricated into individual integrated circuit modules, or multiple modules or steps among them can be fabricated into a single integrated circuit module for implementation. The present invention is not limited to any specific combination of hardware and software.

[0086] Although the specific implementation manners of the present invention have been described above in conjunction with the accompanying drawings, it is not a limitation to the protection scope of the present invention. Those skilled in the art should understand that, based on the technical solutions of the present invention, various modifications or deformations that can be made without creative efforts by those skilled in the art are still within the protection scope of the present invention.

Claims

1. A reliability evaluation method for a battery thermal management system, characterized in that Including: Obtain the discharge timing data and real-time charging temperature timing data of the temperature measurement points of the battery thermal management system at different stages. Among them, the discharge timing data at different stages includes: real-time discharge temperature timing data, SOC timing data, initial cycle discharge temperature timing data, SOC timing data; and temperature difference timing data and SOC difference timing data of the discharge operating conditions adjacent to the real-time operating conditions before the cycle. Input the discharge timing data and the real-time charging temperature timing data into the trained heat dissipation performance prediction model and prediction result verification model respectively to obtain the discharge temperature prediction result and the charging temperature prediction result; evaluate the reliability of the battery thermal management system based on the prediction results. Among them, the heat dissipation performance prediction model includes parallel encoders. The parallel encoders are used to extract features from the discharge timing data at different stages respectively, and the extracted features are fused and then decoded to obtain the discharge temperature prediction result.

2. The reliability evaluation method of a battery thermal management system according to claim 1, wherein The heat dissipation performance prediction model adopts an improved Transformer model. The improved Transformer model includes juxtaposed encoders, and the juxtaposed encoders are respectively an upper limit correction encoder, a main encoder and a performance attenuation encoder. Use the upper limit correction encoder to extract features from the initial cycle discharge temperature timing data and SOC timing data to obtain the first feature. Use the main encoder to extract features from the real-time discharge temperature timing data and SOC timing data to obtain the second feature. Use the performance attenuation encoder to extract features from the temperature difference timing data and SOC difference timing data of the discharge operating conditions adjacent to the real-time operating conditions before the cycle to obtain the third feature. Fuse the first feature, the second feature and the third feature and then decode to obtain the discharge temperature prediction result.

3. The reliability evaluation method of a battery thermal management system according to claim 2, characterized in that Fusing the first feature, the second feature and the third feature and then decoding to obtain the discharge temperature prediction result specifically includes: Construct the weight coefficients of the first feature, the second feature and the third feature respectively according to the reference value of the battery life characterized by the number of cycles and the total number of charge times. Multiply the weight coefficients by the corresponding features and then add them to obtain the fused feature. Based on the fused feature, predict through the discharge condition decoder to obtain the discharge temperature prediction result.

4. The reliability evaluation method of a battery thermal management system according to claim 3, characterized in that The calculation of the weight coefficient is specifically: Where w1 is the weight coefficient corresponding to the first feature, w2 is the weight coefficient corresponding to the second feature, w3 is the weight coefficient corresponding to the third feature, q is the battery life reference value characterized by the number of cycles, and l is the total number of charge times of the battery up to now.

5. A reliability evaluation method for a battery thermal management system according to claim 1, characterized in that Use training samples to train the heat dissipation performance prediction model. The training samples include the temperature timing data and SOC timing data of each temperature point in the initial cycle batch, and the temperature timing data and SOC timing data of each temperature point in the second cycle batch within the sliding window. The temperature difference timing data and SOC difference timing data of each temperature point in adjacent cycle batches.

6. The reliability evaluation method of a battery thermal management system according to claim 5, characterized in that, For the temperature data at the same temperature measurement point in the temperature data of n cyclic batch discharge conditions, the temperature difference sequence between cyclic batches is obtained by subtracting them batch by batch for adjacent cycles, numbered in sequence according to the subtraction batches, and synthesized with the time dimension to obtain the corresponding temperature time series data.

7. The reliability evaluation method of a battery thermal management system according to claim 1, characterized in that Based on the prediction results, evaluate the reliability of the battery thermal management system, specifically: If the predicted discharge temperature prediction result value is higher than the measured temperature under subsequent discharge conditions, it is judged that the reliability of the battery thermal management system is good, otherwise it is judged that the reliability is reduced; If the predicted charging temperature prediction result value is higher than the measured temperature under subsequent charging conditions, it is judged that the reliability of the battery thermal management system is good, otherwise it is judged that the reliability is reduced; When the reliability conclusions judged according to the discharge temperature prediction result and the charging temperature prediction result are different, the result judged according to the charging temperature prediction result is used as the final reliability evaluation result.

8. A reliability evaluation system for a battery thermal management system, characterized in that, Including: An acquisition module for acquiring the discharge time series data and real-time charging temperature time series data at different stages of the temperature measurement points of the battery thermal management system. Among them, the discharge time series data at different stages include: real-time discharge temperature time series data, SOC time series data, initial cycle discharge temperature time series data, SOC time series data; and temperature difference time series data and SOC difference time series data of the cyclic discharge conditions adjacent to before the real-time operation condition. An evaluation module for inputting the discharge time series data and the real-time charging temperature time series data into a trained heat dissipation performance prediction model and a prediction result verification model respectively to obtain a discharge temperature prediction result and a charging temperature prediction result; evaluate the reliability of the battery thermal management system based on the prediction results. Among them, the heat dissipation performance prediction model includes parallel encoders, and the parallel encoders are used to extract features from the discharge time series data at different stages respectively, and the extracted features are fused and then decoded to obtain the discharge temperature prediction result.

9. A computer device, characterized in that, Including: A processor, a memory and a bus. The memory stores machine-readable instructions executable by the processor. When the computer device runs, the processor communicates with the memory through the bus. When the machine-readable instructions are executed by the processor, it executes a method for evaluating the reliability of a battery thermal management system according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, A computer program is stored on the computer-readable storage medium, and when the computer program is run by the processor, it executes a method for evaluating the reliability of a battery thermal management system according to any one of claims 1 to 7.

Images