A method for focused control of battery thermal runaway

Abstract

The present application provides a method for focused control of battery thermal runaway, which relates to the technical field of battery state monitoring. The method includes: performing real-time monitoring on a target battery compartment to obtain a battery compartment monitoring data set, performing standardization processing and feature recognition, building a battery compartment thermal runaway prediction module, then performing thermal runaway prediction to obtain a comprehensive thermal runaway prediction coefficient, determining whether it is less than the thermal runaway threshold, and if it is greater than or equal to the predetermined thermal runaway threshold, generating a thermal runaway warning signal and performing thermal runaway operation and maintenance on the target battery compartment. Through the present application, the technical problem in the prior art that the means for monitoring the state of the battery compartment are single, resulting in inaccurate identification of the thermal runaway risk and further affecting the safety and operation and maintenance efficiency of the battery system can be solved, and the safety and reliability of the battery system are improved.

Description

Technical Field

[0001] This application relates to the technical field of battery state monitoring, and particularly to a method for focused control of battery thermal runaway. Background Art

[0002] The wide application of lithium-ion batteries: Due to their advantages such as high energy density, no radiation, low self-discharge, and low memory effect, with the improvement of battery energy density and charging speed, the safety issues of lithium-ion batteries have become increasingly prominent. Research shows that when batteries are subjected to abusive conditions, safety problems such as thermal runaway may occur. The characteristic of thermal runaway is the generation of intense heat inside the battery, and it may release high-temperature combustible flue gas, releasing a large amount of energy through a chain reaction of chemical components. In a battery module, once a certain battery undergoes thermal runaway, the heat will quickly spread to adjacent batteries, leading to the spread of thermal runaway, which may further trigger more serious safety accidents.

[0003] If only relying on a single monitoring means, such as only monitoring through temperature sensors or gas sensors, it may not be possible to comprehensively capture the complex changes in the battery compartment. This limitation will lead to inaccurate or delayed identification of the thermal runaway risk, thus preventing timely adoption of effective preventive and control measures before a thermal runaway event occurs. This inaccuracy not only increases the risk of accidents in the battery system but may also affect the operation and maintenance efficiency of the entire system, because false warnings or missed reports will result in unnecessary downtime, maintenance, or incorrect operations.

[0004] In summary, the existing technology has a single means of monitoring the battery compartment state, resulting in inaccurate identification of the thermal runaway risk, further affecting the safety and operation and maintenance efficiency of the battery system. Summary of the Invention

[0005] The purpose of this application is to provide a method for focused control of battery thermal runaway to solve the technical problem that the existing technology has a single means of monitoring the battery compartment state, resulting in inaccurate identification of the thermal runaway risk, further affecting the safety and operation and maintenance efficiency of the battery system.

[0006] In view of the above problems, this application provides a method for focused control of battery thermal runaway.

[0007] The present application provides a method for focused control of battery thermal runaway. The method includes: obtaining a battery compartment monitoring data set by monitoring a target battery compartment in real time according to a battery compartment monitoring unit; performing normalization processing and feature recognition on the battery compartment monitoring data set to generate battery compartment temperature feature data, battery compartment gas feature data, and battery compartment operation feature data; building a battery compartment thermal runaway prediction module, where the battery compartment thermal runaway prediction module includes a first thermal runaway prediction unit, a second thermal runaway prediction unit, a third thermal runaway prediction unit, and a thermal runaway fusion prediction unit; performing thermal runaway prediction according to the battery compartment thermal runaway prediction module, combining the battery compartment temperature feature data, the battery compartment gas feature data, and the battery compartment operation feature data to obtain a comprehensive thermal runaway prediction coefficient; determining whether the comprehensive thermal runaway prediction coefficient is less than a predetermined thermal runaway threshold; if the comprehensive thermal runaway prediction coefficient is greater than or equal to the predetermined thermal runaway threshold, generating a thermal runaway warning signal; and activating a thermal runaway focused control module according to the thermal runaway warning signal to perform thermal runaway operation and maintenance on the target battery compartment.

[0008] One or more technical solutions provided in the present application have at least the following technical effects or advantages:

[0009] By obtaining a battery compartment monitoring data set by monitoring a target battery compartment in real time according to a battery compartment monitoring unit; performing normalization processing and feature recognition on the battery compartment monitoring data set to generate battery compartment temperature feature data, battery compartment gas feature data, and battery compartment operation feature data; building a battery compartment thermal runaway prediction module, where the battery compartment thermal runaway prediction module includes a first thermal runaway prediction unit, a second thermal runaway prediction unit, a third thermal runaway prediction unit, and a thermal runaway fusion prediction unit; performing thermal runaway prediction according to the battery compartment thermal runaway prediction module, combining the battery compartment temperature feature data, the battery compartment gas feature data, and the battery compartment operation feature data to obtain a comprehensive thermal runaway prediction coefficient; determining whether the comprehensive thermal runaway prediction coefficient is less than a predetermined thermal runaway threshold; if the comprehensive thermal runaway prediction coefficient is greater than or equal to the predetermined thermal runaway threshold, generating a thermal runaway warning signal; and activating a thermal runaway focused control module according to the thermal runaway warning signal to perform thermal runaway operation and maintenance on the target battery compartment, the technical problem in the prior art that the single means of monitoring the state of the battery compartment leads to inaccurate identification of the thermal runaway risk, further affecting the safety and operation and maintenance efficiency of the battery system is effectively solved, and the safety and reliability of the battery system are improved.

[0010] The above description is only an overview of the technical solution of the present application. In order to better understand the technical means of the present application, it can be implemented according to the content of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and understandable, the specific embodiments of the present application are specifically given below. It should be understood that the content described in this part is not intended to identify the key or important features of the embodiments of the present application, nor is it used to limit the scope of the present application. Other features of the present application will become easily understood through the following description. Brief Description of the Drawings

[0011] In order to more clearly illustrate the technical solutions in the present application or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only exemplary. For those of ordinary skill in the art, other drawings can be obtained according to the provided drawings without creative efforts.

[0012] Figure 1 It is a schematic flowchart of a method for controlling battery thermal runaway focusing in the present application;

[0013] Figure 2 It is a schematic flowchart of performing thermal runaway operation and maintenance in a method for controlling battery thermal runaway focusing in the present application. Detailed Embodiments

[0014] By providing a method for controlling battery thermal runaway focusing, the present application solves the technical problems in the prior art that the means for monitoring the state of the battery compartment are single, resulting in inaccurate identification of thermal runaway risks, further affecting the safety and operation and maintenance efficiency of the battery system, and improves the safety and reliability of the battery system.

[0015] Next, the technical solutions in the present application will be clearly and completely described with reference to the drawings. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. It should be understood that the present application is not limited by the exemplary embodiments described here. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application without creative efforts belong to the scope of protection of the present application. In addition, it should be noted that for the sake of description, only the parts related to the present application are shown in the drawings rather than all.

[0016] Embodiment 1

[0017] Please refer to the attached Figure 1 , the present application provides a method for controlling battery thermal runaway focusing. Among them, the method specifically includes the following steps:

[0018] S1: Real-time monitor the target battery compartment according to the battery compartment monitoring unit to obtain a battery compartment monitoring data set;

[0019] Specifically, by means of the real-time monitoring of the target battery compartment by the battery compartment monitoring unit, the following battery compartment monitoring data sets can be obtained, including battery state parameters, voltage. Through real-time monitoring, the voltage values of each battery in the battery compartment can be obtained. Current, monitoring the current of the battery compartment can reflect the charge and discharge state of the battery and whether there is abnormal current. Temperature, the operating temperature of the battery has an important impact on its performance and safety. By real-time monitoring the temperature of the battery compartment, temperature anomalies can be detected in a timely manner to prevent safety problems such as thermal runaway. Environmental parameters, humidity, the humidity in the battery compartment may affect the performance and lifespan of the battery. By monitoring the humidity, humidity anomalies can be detected in a timely manner and corresponding measures can be taken. Air pressure, changes in air pressure may affect the sealing and safety of the battery. Real-time monitoring of air pressure helps to detect potential safety hazards in a timely manner. Each piece of monitoring data will be accompanied by a timestamp recording the exact time of data collection.

[0020] S2: Perform standardization processing and feature recognition based on the battery compartment monitoring data set to generate battery compartment temperature feature data, battery compartment gas feature data, and battery compartment operation feature data;

[0021] Specifically, remove duplicate data, which can be filled by interpolation, deletion, or estimation based on other relevant features. Identify and process outliers, such as using statistical methods like the IQR rule, Z-score, etc. to identify and process extreme data. Normalize or standardize data with different dimensions so that each feature is comparable. Normalization is to scale the data to the range of [0, 1] or [-1, 1]. Standardization is to convert the data into a distribution with a mean of 0 and a standard deviation of 1. For categorical data, such as the identification information of the battery compartment, protection measure activation information, etc., one-hot encoding or label encoding can be used for conversion. Extract data related to temperature from the monitoring data set. Calculate statistical features such as the average value, maximum value, minimum value, standard deviation, etc. of the temperature to generate battery compartment temperature feature data. If the monitoring data set contains gas components or relevant sensor data, such as humidity, air pressure, etc., extract this data. Calculate statistical features such as the concentration, humidity, air pressure of the gas, such as the mean value, variance, etc. to generate battery compartment gas feature data. Analyze the change trend and periodicity of the gas components. Extract data related to the operation state of the battery compartment, such as charge and discharge current, voltage, etc. Calculate statistical features of these operation parameters, such as the average value, peak value, fluctuation range of current and voltage, etc. to generate battery compartment operation feature data. Analyze the time series characteristics of the operation parameters, such as periodic charge and discharge patterns, abnormal operation patterns, etc.

[0022] S3: Build a battery compartment thermal runaway prediction module, where the battery compartment thermal runaway prediction module includes a first thermal runaway prediction unit, a second thermal runaway prediction unit, a third thermal runaway prediction unit, and a thermal runaway fusion prediction unit;

[0023] Specifically, receive and process the temperature characteristic data of the battery compartment, including temperature statistical characteristics, temperature change rate, etc. Use machine learning algorithms such as SVM, random forest, etc. or deep learning models to learn the temperature data and establish an association model between temperature characteristics and thermal runaway. Output the thermal runaway prediction result based on temperature characteristics. Receive and process the gas characteristic data of the battery compartment, such as statistical characteristics and change trends of gas concentration, humidity, air pressure, etc. Adopt models suitable for processing time series data, such as LSTM, GRU, etc., to learn the relationship between gas characteristics and thermal runaway. Output the thermal runaway prediction result based on gas characteristics. Receive and process the operation characteristic data of the battery compartment, including statistical characteristics and time series characteristics of operation parameters such as current and voltage. Use algorithms such as regression analysis and neural network to establish a prediction model between operation characteristics and thermal runaway. Output the thermal runaway prediction result based on operation characteristics. Receive the output results of the first, second, and third thermal runaway prediction units. Adopt weighted average, voting mechanism or fusion algorithms such as stacked ensemble learning to fuse the results of each prediction unit. According to the fusion result, judge whether the battery compartment will have thermal runaway and output the final prediction result. A comprehensive and efficient thermal runaway prediction module for the battery compartment can be constructed to provide strong guarantee for the safe operation of the battery system.

[0024] S4: According to the thermal runaway prediction module of the battery compartment, combine the temperature characteristic data, gas characteristic data, and operation characteristic data of the battery compartment to conduct thermal runaway prediction and obtain a comprehensive thermal runaway prediction coefficient;

[0025] Specifically, input the temperature characteristic data of the battery compartment into the first thermal runaway prediction unit. Input the gas characteristic data of the battery compartment into the second thermal runaway prediction unit. Input the operation characteristic data of the battery compartment into the third thermal runaway prediction unit. Each prediction unit processes the data according to its specific algorithm and model and outputs a predicted value or probability of thermal runaway. The thermal runaway fusion prediction unit receives the outputs of the three prediction units. The fusion prediction unit can adopt weighted average, voting, or more complex fusion algorithms such as weight-based fusion and decision tree fusion to synthesize the results of the three prediction units. The result of the fusion is a comprehensive thermal runaway prediction coefficient, which can be a specific value representing the risk degree of thermal runaway or a classification label such as high risk, medium risk, or low risk. Finally, the thermal runaway fusion prediction unit will output a comprehensive thermal runaway prediction coefficient, which represents a comprehensive assessment of the thermal runaway risk of the battery compartment based on all available characteristic data currently.

[0026] S5: Judge whether the comprehensive thermal runaway prediction coefficient is less than a predetermined thermal runaway threshold;

[0027] Specifically, this threshold can be set based on historical data. The threshold can be a fixed value or a value that is dynamically adjusted according to the status of the battery compartment. The comprehensive thermal runaway prediction coefficient is directly compared with the predetermined thermal runaway threshold. If the comprehensive thermal runaway prediction coefficient is less than the predetermined thermal runaway threshold, then the thermal runaway risk of the current battery compartment is considered acceptable. If the comprehensive thermal runaway prediction coefficient is greater than or equal to the predetermined thermal runaway threshold, then it means that there is a relatively high thermal runaway risk in the battery compartment.

[0028] S6: If the comprehensive thermal runaway prediction coefficient is greater than or equal to the predetermined thermal runaway threshold, generate a thermal runaway warning signal;

[0029] Specifically, continuously monitor the comprehensive thermal runaway prediction coefficient and compare it with the predetermined thermal runaway threshold. Once it is found that the comprehensive thermal runaway prediction coefficient is greater than or equal to the predetermined thermal runaway threshold, the triggering condition is satisfied. When the triggering condition is satisfied, a thermal runaway warning signal is automatically generated. This warning signal can be an electronic alarm, an audible alarm, a light indication, or a prominent mark displayed on the monitoring interface.

[0030] S7: Activate the thermal runaway focusing control module according to the thermal runaway warning signal to perform thermal runaway operation and maintenance on the target battery compartment.

[0031] Specifically, the thermal runaway focusing control module first receives the thermal runaway warning signal and identifies the level and source of the signal. According to the severity of the warning signal, determine the urgency of the operation and maintenance and the measures to be taken. According to the pre-developed thermal runaway emergency plan, the control module will automatically or prompt the operator to manually initiate a series of emergency measures. These measures include cutting off the power supply, starting the emergency cooling system, isolating the battery compartment, etc. The control module will strengthen the real-time monitoring of the target battery compartment, including tracking key parameters such as temperature, pressure, and gas composition. Conduct real-time analysis on the collected data to evaluate the risk and development trend of thermal runaway. According to the data analysis results, the control module will automatically adjust the environmental parameters inside the battery compartment, such as increasing the ventilation volume, lowering the temperature, etc., to reduce the risk of thermal runaway.

[0032] Further, step S3 of the present application further includes:

[0033] Interact with the battery compartment data management unit to construct the first thermal runaway prediction unit;

[0034] Interact with the battery compartment data management unit to construct the second thermal runaway prediction unit;

[0035] Interact with the battery compartment data management unit to construct the third thermal runaway prediction unit;

[0036] The thermal runaway fusion prediction unit is established, wherein the thermal runaway fusion prediction unit includes a thermal runaway fusion prediction function, and the thermal runaway fusion prediction function is:

[0037] ;

[0038] wherein, TLCD represents the comprehensive thermal runaway prediction coefficient, TLCA represents the first thermal runaway prediction coefficient, TLCB represents the second thermal runaway prediction coefficient, and TLCC represents the third thermal runaway prediction coefficient;

[0039] The first thermal runaway prediction unit, the second thermal runaway prediction unit, the third thermal runaway prediction unit and the thermal runaway fusion prediction unit are fused to obtain the battery compartment thermal runaway prediction module.

[0040] Specifically, temperature characteristic data is obtained from the battery compartment data management unit. Based on these data, a model is trained using machine learning or deep learning algorithms to construct the first thermal runaway prediction unit, which can output a thermal runaway prediction coefficient based on temperature characteristics. Gas characteristic data is obtained from the battery compartment data management unit. Using these data to train a model to construct the second thermal runaway prediction unit, which can output a thermal runaway prediction coefficient based on gas characteristics. Interact with the battery compartment data management unit to construct the third thermal runaway prediction unit: obtain operation characteristic data from the battery compartment data management unit. Using these data to train a model to construct the third thermal runaway prediction unit, which can output a thermal runaway prediction coefficient based on operation characteristics. Create a fusion prediction unit that will receive the outputs from the first three prediction units. In the fusion prediction unit, a thermal runaway fusion prediction function is defined to calculate the comprehensive thermal runaway prediction coefficient. The specific function is, ; wherein, TLCD represents the comprehensive thermal runaway prediction coefficient, TLCA represents the first thermal runaway prediction coefficient, TLCB represents the second thermal runaway prediction coefficient, and TLCC represents the third thermal runaway prediction coefficient; input the outputs of the first, second, and third thermal runaway prediction units into the thermal runaway fusion prediction unit. Apply the thermal runaway fusion prediction function to calculate the comprehensive thermal runaway prediction coefficient. By integrating three prediction units and a fusion prediction unit, a complete battery compartment thermal runaway prediction module is obtained. This module can perform a comprehensive thermal runaway risk assessment based on multiple characteristic data and output a comprehensive thermal runaway prediction coefficient.

[0041] Furthermore, this application also includes:

[0042] According to the battery compartment data management unit, retrieve the battery compartment temperature characteristic data record, the battery compartment temperature anomaly detection record, and the temperature characteristic thermal runaway coefficient record;

[0043] Using the battery compartment temperature characteristic data record as input data and the battery compartment temperature anomaly detection record as output data, train a predetermined machine learning model. Each time after training a predetermined number of times, obtain the mean squared loss error of temperature anomaly detection;

[0044] If the mean squared loss error of temperature anomaly detection is less than or equal to a predetermined mean squared loss error threshold, generate a battery compartment temperature anomaly detection model;

[0045] Using the battery compartment temperature anomaly detection record as input information and the temperature characteristic thermal runaway coefficient record as output information, train the predetermined machine learning model. Each time after training the predetermined number of times, obtain the mean squared loss error of thermal runaway prediction;

[0046] If the mean squared loss error of thermal runaway prediction is less than or equal to the predetermined mean squared loss error threshold, generate a temperature characteristic thermal runaway prediction model;

[0047] Fully connect the battery compartment temperature anomaly detection model and the temperature characteristic thermal runaway prediction model to generate the first thermal runaway prediction unit.

[0048] Specifically, obtain the battery compartment temperature characteristic data record, which will be used as the input of the machine learning model. Obtain the battery compartment temperature anomaly detection record, which will be used as the target output when training the temperature anomaly detection model. Obtain the temperature characteristic thermal runaway coefficient record, which will be used as the target output when training the thermal runaway prediction model. Use machine learning algorithms such as support vector machines, random forests, or neural networks. Use the battery compartment temperature characteristic data record as input data. Use the battery compartment temperature anomaly detection record as output data. And after each training of a predetermined number of times, calculate the mean squared loss error of temperature anomaly detection. If the mean squared loss error of temperature anomaly detection is less than or equal to the predetermined mean squared loss error threshold, it indicates that the model performance meets the requirements. Generate a battery compartment temperature anomaly detection model, which can detect the temperature anomaly of the battery compartment based on the temperature characteristic data. Use the same machine learning algorithm, use the battery compartment temperature anomaly detection record as input information. Use the temperature characteristic thermal runaway coefficient record as output information. Start training the model, and after each training of a predetermined number of times, obtain the mean squared loss error of thermal runaway prediction. If the mean squared loss error of thermal runaway prediction is less than or equal to the predetermined mean squared loss error threshold, it indicates that the model performance meets the standard. Generate a temperature characteristic thermal runaway prediction model, which can predict the thermal runaway coefficient based on the temperature anomaly detection record. Fully connect the battery compartment temperature anomaly detection model and the temperature characteristic thermal runaway prediction model. The output of the temperature anomaly detection model will be used as the input of the thermal runaway prediction model. By fully connecting the two trained models, the first thermal runaway prediction unit is obtained. This unit can receive the temperature characteristic data of the battery compartment, first detect the temperature anomaly, and then predict the thermal runaway coefficient based on these anomalies.

[0049] Further, step S4 of the present application further includes:

[0050] Input the temperature characteristic data of the battery compartment into the first thermal runaway prediction unit to obtain a temperature characteristic thermal runaway prediction coefficient;

[0051] Input the gas characteristic data of the battery compartment into the second thermal runaway prediction unit to obtain a gas characteristic thermal runaway prediction coefficient;

[0052] Input the operation characteristic data of the battery compartment into the third thermal runaway prediction unit to obtain an operation characteristic thermal runaway prediction coefficient;

[0053] Perform normalization processing on the temperature characteristic thermal runaway prediction coefficient, the gas characteristic thermal runaway prediction coefficient, and the operation characteristic thermal runaway prediction coefficient to obtain a first thermal runaway prediction coefficient, a second thermal runaway prediction coefficient, and a third thermal runaway prediction coefficient;

[0054] Input the first thermal runaway prediction coefficient, the second thermal runaway prediction coefficient, and the third thermal runaway prediction coefficient into the thermal runaway fusion prediction unit to output the comprehensive thermal runaway prediction coefficient.

[0055] Specifically, input the temperature characteristic data of the battery compartment into the first thermal runaway prediction unit. Through the processing of this unit, a temperature characteristic thermal runaway prediction coefficient is obtained. Input the gas characteristic data of the battery compartment into the second thermal runaway prediction unit. Through the processing of this unit, a gas characteristic thermal runaway prediction coefficient is obtained. Input the operation characteristic data of the battery compartment into the third thermal runaway prediction unit. Through the processing of this unit, an operation characteristic thermal runaway prediction coefficient is obtained. Perform normalization processing on the temperature characteristic thermal runaway prediction coefficient, the gas characteristic thermal runaway prediction coefficient, and the operation characteristic thermal runaway prediction coefficient. Normalization processing can eliminate the dimensional difference and numerical range difference between different characteristic data, enabling each prediction coefficient to be fused on the same scale. The coefficients after normalization processing are respectively called the first thermal runaway prediction coefficient, the second thermal runaway prediction coefficient, and the third thermal runaway prediction coefficient. Input the first thermal runaway prediction coefficient, the second thermal runaway prediction coefficient, and the third thermal runaway prediction coefficient after normalization processing into the thermal runaway fusion prediction unit. The thermal runaway fusion prediction unit performs fusion processing on these coefficients according to a preset fusion prediction function. Through the processing of the thermal runaway fusion prediction unit, a comprehensive thermal runaway prediction coefficient is finally output.

[0056] Further, the present application further includes:

[0057] Input the gas characteristic data of the battery compartment into the battery compartment gas anomaly detection model in the second thermal runaway prediction unit to generate gas anomaly detection characteristic data;

[0058] Input the gas anomaly detection feature data into the gas feature thermal runaway prediction model in the second thermal runaway prediction unit, and output the gas feature thermal runaway prediction coefficient.

[0059] Specifically, input the gas feature data of the battery compartment into the gas anomaly detection model of the battery compartment in the second thermal runaway prediction unit. The model has been trained to identify abnormal situations that do not conform to the normal gas feature pattern, such as sudden increases or decreases in gas concentration, or the abnormal appearance of a certain specific gas. The gas anomaly detection model of the battery compartment processes the input gas feature data and generates gas anomaly detection feature data. These feature data include information such as the detected abnormal points, the degree of abnormality, and the type of abnormality, which are crucial for subsequent thermal runaway prediction. Next, input the gas anomaly detection feature data into the gas feature thermal runaway prediction model in the second thermal runaway prediction unit. The model has been trained to predict the risk of thermal runaway in the battery compartment based on the abnormal feature data. The gas feature thermal runaway prediction model processes the input gas anomaly detection feature data and outputs the gas feature thermal runaway prediction coefficient. This coefficient represents the likelihood or risk level of thermal runaway in the battery compartment based on the current gas feature data and the detected abnormal situations.

[0060] Further, as Figure 2 shown, step S7 of this application further includes:

[0061] The thermal runaway focusing control module includes a thermal runaway focusing unit, a thermal runaway operation and maintenance decision-making unit, and a thermal runaway regulation unit;

[0062] Perform thermal runaway feature focusing according to the thermal runaway focusing unit to generate a thermal runaway feature focusing result;

[0063] Based on the thermal runaway feature focusing result, combine with the thermal runaway operation and maintenance decision-making unit to obtain a thermal runaway operation and maintenance feature decision;

[0064] Transmit the thermal runaway operation and maintenance feature decision to the thermal runaway regulation unit, and the thermal runaway regulation unit performs the thermal runaway operation and maintenance of the target battery compartment based on the thermal runaway operation and maintenance feature decision.

[0065] Specifically, the thermal runaway focusing unit is responsible for analyzing and focusing on the characteristics related to thermal runaway. This includes collecting data from various sensors and monitoring systems, identifying key parameters and trends related to the risk of thermal runaway. Through data analysis and feature extraction techniques, a thermal runaway feature focusing result is generated. This result is a set of one or more key indicators used to indicate the potential risk and urgency of thermal runaway. The thermal runaway operation and maintenance decision-making unit receives the feature focusing result from the thermal runaway focusing unit. Based on these results, combined with preset operation and maintenance strategies and historical data, this unit formulates one or more thermal runaway operation and maintenance feature decisions. These decisions include emergency cooling measures, isolating potentially dangerous areas, notifying relevant personnel, etc. The thermal runaway regulation unit is responsible for receiving and executing the instructions from the thermal runaway operation and maintenance decision-making unit. According to the operation and maintenance feature decisions, the regulation unit will trigger corresponding control measures, such as starting the cooling system, shutting down the power supply in a specific area, issuing an alarm, etc.

[0066] Furthermore, this application also includes:

[0067] Retrieve multi-dimensional thermal runaway feature data according to the battery compartment thermal runaway prediction module;

[0068] Based on the multi-dimensional thermal runaway feature data, perform thermal runaway fitting propagation on the multi-dimensional thermal runaway feature data according to the thermal runaway propagation model in the thermal runaway focusing unit to generate fitting thermal runaway propagation feature data;

[0069] Perform data fusion based on the multi-dimensional thermal runaway feature data and the fitting thermal runaway propagation feature data to obtain the thermal runaway feature focusing result.

[0070] Specifically, multi-dimensional thermal runaway characteristic data is retrieved from the battery compartment thermal runaway prediction module. These data include, but are not limited to, various parameters such as temperature, pressure, gas composition, current, and voltage, which reflect the thermal runaway risk of the battery compartment from different dimensions. The retrieved multi-dimensional thermal runaway characteristic data is input into the thermal runaway propagation model within the thermal runaway focusing unit. The thermal runaway propagation model is a trained mathematical model used to simulate and predict the propagation behavior of thermal runaway within the battery compartment. The multi-dimensional thermal runaway characteristic data is fitted using the thermal runaway propagation model to simulate the possible propagation paths and speeds of thermal runaway within the battery compartment. Through the fitting process, fitted thermal runaway propagation characteristic data is generated, which reflects the propagation characteristics of thermal runaway in terms of time and space. The original multi-dimensional thermal runaway characteristic data and the fitted thermal runaway propagation characteristic data are prepared for data fusion. Appropriate data fusion techniques, such as weighted average, Kalman filtering, neural network fusion, etc., are used to fuse the original characteristic data and the fitted propagation characteristic data. The purpose of fusion is to comprehensively utilize the information from both data sources to improve the accuracy and integrity of the description of thermal runaway characteristics. After data fusion, a thermal runaway characteristic focusing result is obtained. This result is a dataset that combines multi-dimensional original characteristics and fitted propagation characteristics, and can more comprehensively and accurately describe the current thermal runaway state and trend of the battery compartment.

[0071] Furthermore, this application also includes:

[0072] Obtain the multi-dimensional thermal runaway characteristic data record library of the target battery compartment;

[0073] Perform temporal processing based on the multi-dimensional thermal runaway characteristic data record library to generate a multi-dimensional thermal runaway characteristic record sequence;

[0074] Load the extended thermal runaway characteristic data record library of the target battery compartment, and perform temporal processing on the extended thermal runaway characteristic data record library to obtain a multi-dimensional extended thermal runaway characteristic record sequence;

[0075] Supervise and train the Markov chain based on the multi-dimensional extended thermal runaway characteristic record sequence to obtain an initial thermal runaway propagation model;

[0076] Perform incremental learning on the initial thermal runaway propagation model based on the multi-dimensional thermal runaway characteristic record sequence to generate the thermal runaway propagation model.

[0077] Specifically, obtain the multi-dimensional thermal runaway characteristic data repository of the target battery compartment. This repository may contain various characteristic data related to thermal runaway, such as temperature, pressure, current, voltage, gas composition, etc. Perform time-series processing on the multi-dimensional thermal runaway characteristic data repository. Arrange the data in chronological order to generate a multi-dimensional thermal runaway characteristic record sequence. Load the extended thermal runaway characteristic data repository of the target battery compartment. This extended repository contains more thermal runaway data in more scenarios and of more types, which is used to enhance the generalization ability of the model. Also perform time-series processing on the extended thermal runaway characteristic data repository to obtain a multi-dimensional extended thermal runaway characteristic record sequence. Use the multi-dimensional extended thermal runaway characteristic record sequence to conduct supervised training on the Markov chain. The Markov chain is a stochastic process where the future state depends only on the current state, and this model is suitable for describing phenomena with time-series dependence such as thermal runaway. Through supervised training, obtain an initial thermal runaway propagation model. This model can predict the future thermal runaway state based on the current state. Use the multi-dimensional thermal runaway characteristic record sequence to perform incremental learning on the initial thermal runaway propagation model. Incremental learning allows the model to continuously learn and improve while continuously receiving new data, without the need to retrain the entire model. Through incremental learning, the model can gradually adapt to the unique thermal runaway characteristics of the target battery compartment, improving the accuracy and reliability of the prediction. After incremental learning, generate the final thermal runaway propagation model. This model combines the extended data and the specific data of the target battery compartment, and can more accurately simulate and predict the propagation process of thermal runaway.

[0078] In summary, a battery thermal runaway focusing control method provided by this application has the following technical effects:

[0079] By performing real-time monitoring on a target battery compartment according to a battery compartment monitoring unit, a battery compartment monitoring data set is obtained; standardization processing and feature recognition are performed on the battery compartment monitoring data set to generate battery compartment temperature feature data, battery compartment gas feature data, and battery compartment operation feature data; a battery compartment thermal runaway prediction module is built, wherein the battery compartment thermal runaway prediction module includes a first thermal runaway prediction unit, a second thermal runaway prediction unit, a third thermal runaway prediction unit, and a thermal runaway fusion prediction unit; according to the battery compartment thermal runaway prediction module, combining the battery compartment temperature feature data, the battery compartment gas feature data, and the battery compartment operation feature data to perform thermal runaway prediction, and a comprehensive thermal runaway prediction coefficient is obtained; it is judged whether the comprehensive thermal runaway prediction coefficient is less than a predetermined thermal runaway threshold; if the comprehensive thermal runaway prediction coefficient is greater than or equal to the predetermined thermal runaway threshold, a thermal runaway warning signal is generated; according to the thermal runaway warning signal, a thermal runaway focusing control module is activated to perform thermal runaway operation and maintenance on the target battery compartment, effectively solving the technical problem in the prior art that the means for monitoring the state of the battery compartment is single, resulting in inaccurate identification of the thermal runaway risk, and further affecting the safety and operation and maintenance efficiency of the battery system, and improving the safety and reliability of the battery system.

[0080] The foregoing description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Thus, the present application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0081] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalent technologies, the present application is also intended to include these changes and modifications.

Claims

1. A battery thermal runaway focusing control method, characterized in that: The method comprises: Perform real-time monitoring of the target battery compartment according to the battery compartment monitoring unit to obtain a battery compartment monitoring data set; Performing standardization processing and feature recognition on the battery compartment monitoring data set to generate battery compartment temperature feature data, battery compartment gas feature data, and battery compartment operation feature data; Building a battery compartment thermal runaway prediction module, wherein the battery compartment thermal runaway prediction module includes a first thermal runaway prediction unit, a second thermal runaway prediction unit, a third thermal runaway prediction unit and a thermal runaway fusion prediction unit; According to the battery compartment thermal runaway prediction module, thermal runaway prediction is performed in combination with the battery compartment temperature characteristic data, the battery compartment gas characteristic data and the battery compartment operation characteristic data to obtain a comprehensive thermal runaway prediction coefficient; Determining whether the comprehensive thermal runaway prediction coefficient is less than a predetermined thermal runaway threshold; If the comprehensive thermal runaway prediction coefficient is greater than / equal to the predetermined thermal runaway threshold, generating a thermal runaway warning signal; activating a thermal runaway focus control module according to the thermal runaway warning signal to perform thermal runaway operation and maintenance on the target battery compartment; Build a battery compartment thermal runaway prediction module, including: Interacting with the battery compartment data management unit to construct the first thermal runaway prediction unit; Interacting with the battery compartment data management unit to construct the second thermal runaway prediction unit; Interacting with the battery compartment data management unit to construct the third thermal runaway prediction unit; The thermal runaway fusion prediction unit is established, wherein the thermal runaway fusion prediction unit includes a thermal runaway fusion prediction function, and the thermal runaway fusion prediction function is: ； Among them, TLCD represents the comprehensive thermal runaway prediction coefficient, TLCA represents the first thermal runaway prediction coefficient, TLCB represents the second thermal runaway prediction coefficient, and TLCC represents the third thermal runaway prediction coefficient; The battery compartment thermal runaway prediction module is obtained by fusing the first thermal runaway prediction unit, the second thermal runaway prediction unit, the third thermal runaway prediction unit and the thermal runaway fusion prediction unit.

2. The method according to claim 1, characterized in that The interactive battery compartment data management unit constructs the first thermal runaway prediction unit, including: According to the battery compartment data management unit, retrieve the battery compartment temperature characteristic data record, the battery compartment temperature abnormality detection record and the temperature characteristic thermal runaway coefficient record; Using the battery compartment temperature characteristic data record as input data and the battery compartment temperature anomaly detection record as output data, a predetermined machine learning model is trained, and after each predetermined number of trainings, a mean square loss error of temperature anomaly detection is obtained; If the temperature anomaly detection mean square loss error is less than or equal to a predetermined mean square loss error threshold, generating a battery compartment temperature anomaly detection model; Taking the battery compartment temperature anomaly detection record as input information and the temperature characteristic thermal runaway coefficient record as output information, training the predetermined machine learning model, and obtaining a thermal runaway prediction mean square loss error each time the predetermined number of trainings is performed; If the thermal runaway prediction mean square loss error is less than or equal to the predetermined mean square loss error threshold, generating a temperature characteristic thermal runaway prediction model; The battery compartment temperature anomaly detection model and the temperature characteristic thermal runaway prediction model are fully connected to generate the first thermal runaway prediction unit.

3. The method according to claim 1, characterized in that According to the battery compartment thermal runaway prediction module, thermal runaway prediction is performed in combination with the battery compartment temperature characteristic data, the battery compartment gas characteristic data and the battery compartment operation characteristic data to obtain a comprehensive thermal runaway prediction coefficient, including: Inputting the battery compartment temperature characteristic data into the first thermal runaway prediction unit to obtain a temperature characteristic thermal runaway prediction coefficient; Inputting the battery compartment gas characteristic data into the second thermal runaway prediction unit to obtain a gas characteristic thermal runaway prediction coefficient; Inputting the battery compartment operation characteristic data into the third thermal runaway prediction unit to obtain an operation characteristic thermal runaway prediction coefficient; Performing standardization processing according to the temperature characteristic thermal runaway prediction coefficient, the gas characteristic thermal runaway prediction coefficient and the operation characteristic thermal runaway prediction coefficient to obtain a first thermal runaway prediction coefficient, a second thermal runaway prediction coefficient and a third thermal runaway prediction coefficient; The first thermal runaway prediction coefficient, the second thermal runaway prediction coefficient and the third thermal runaway prediction coefficient are input into the thermal runaway fusion prediction unit, and the comprehensive thermal runaway prediction coefficient is output.

4. The method according to claim 3, characterized in that Inputting the battery compartment gas characteristic data into the second thermal runaway prediction unit to obtain a gas characteristic thermal runaway prediction coefficient includes: inputting the battery compartment gas characteristic data into a battery compartment gas anomaly detection model in the second thermal runaway prediction unit to generate gas anomaly detection characteristic data; The gas anomaly detection characteristic data is input into the gas characteristic thermal runaway prediction model in the second thermal runaway prediction unit, and the gas characteristic thermal runaway prediction coefficient is output.

5. The method according to claim 1, characterized in that Activating a thermal runaway focus control module according to the thermal runaway warning signal to perform thermal runaway operation and maintenance on the target battery compartment includes: The thermal runaway focusing control module includes a thermal runaway focusing unit, a thermal runaway operation and maintenance decision unit, and a thermal runaway control unit; Performing thermal runaway feature focusing according to the thermal runaway focusing unit to generate a thermal runaway feature focusing result; According to the thermal runaway feature focusing result, in combination with the thermal runaway operation and maintenance decision unit, a thermal runaway operation and maintenance feature decision is obtained; The thermal runaway operation and maintenance feature decision is transmitted to the thermal runaway control unit, and the thermal runaway control unit performs thermal runaway operation and maintenance of the target battery compartment based on the thermal runaway operation and maintenance feature decision.

6. The method according to claim 5, characterized in that Performing thermal runaway feature focusing according to the thermal runaway focusing unit to generate a thermal runaway feature focusing result includes: According to the battery compartment thermal runaway prediction module, multi-dimensional thermal runaway characteristic data is retrieved; Based on the multi-dimensional thermal runaway characteristic data, performing thermal runaway fitting propagation on the multi-dimensional thermal runaway characteristic data according to a thermal runaway propagation model in the thermal runaway focusing unit to generate fitted thermal runaway propagation characteristic data; Data fusion is performed based on the multi-dimensional thermal runaway characteristic data and the fitted thermal runaway propagation characteristic data to obtain the thermal runaway characteristic focusing result.

7. The method according to claim 6, characterized in that The method comprises: Obtaining a multi-dimensional thermal runaway characteristic data record library of the target battery compartment; Performing time series processing according to the multi-dimensional thermal runaway characteristic data record library to generate a multi-dimensional thermal runaway characteristic record sequence; Loading the expanded thermal runaway feature data record library of the target battery compartment, and performing time-sequencing processing on the expanded thermal runaway feature data record library to obtain a multi-dimensional expanded thermal runaway feature record sequence; Performing supervised training on the Markov chain according to the multi-dimensional expanded thermal runaway feature record sequence to obtain an initial thermal runaway propagation model; The initial thermal runaway propagation model is incrementally learned according to the multi-dimensional thermal runaway feature record sequence to generate the thermal runaway propagation model.

Images