Lithium ion battery thermal runaway early warning method based on sound signals

Abstract

The invention relates to a lithium ion battery thermal runaway early warning method based on sound signals, and belongs to the technical field of batteries, and the method comprises the following steps: S1, collecting a lithium ion battery thermal runaway valve opening sound data set and an environment noise data set, and establishing a battery sound database; s2, extracting a time domain, a frequency domain and statistical features of the sound data by using a time sequence feature extraction library TSFEL; s3, the correlation of the extracted features is analyzed by adopting a grey correlation gradient GRG, and redundant features with high correlation are eliminated; s4, training a random forest classifier based on a Bayesian optimization algorithm by taking a corresponding sound type as a training label according to the sample data after the features are screened, and constructing a battery thermal runaway early warning model based on sound signals; and S5, taking the test data after feature extraction and selection as the input of a battery thermal runaway early warning model, and judging whether the battery has a thermal runaway risk according to whether a valve opening sound is detected.

Description

Technical Field

[0001] The present invention belongs to the technical field of batteries and relates to a lithium-ion battery thermal runaway early warning method based on sound signals. Background Art

[0002] Accurate prediction and early warning of thermal runaway are fundamental to ensuring the efficient, reliable, and safe operation of new energy vehicles. During thermal runaway, as the temperature rises, the battery undergoes the following processes: decomposition of the solid electrolyte interface, reaction between the electrolyte and the anode active material, melting of the separator, thermal decomposition of the cathode material, redox reactions of the electrolyte, decomposition of the binder, and combustion of the electrolyte. During this period, internal short circuits caused by separator collapse can also contribute to the thermal runaway process. Thermal runaway in lithium-ion batteries is a violent and rapid process, accompanied by the release of large amounts of heat and flammable gases, creating a chain reaction that accelerates the progression of thermal runaway. In electric vehicle battery systems, thermal runaway in a single cell can rapidly spread throughout the entire battery module and even the entire battery system, causing thermal runaway to propagate, leading to a chain reaction and explosion. The explosion of gases released during thermal runaway in lithium-ion batteries is a major cause of lithium-ion battery accidents. Violent explosions can cause serious casualties and property damage. Furthermore, the toxic gas mixture produced by the series of combustion reactions can be released into the air, causing serious environmental problems. The operating conditions of lithium-ion batteries are extremely complex, and thermal runaway cannot be avoided from a mechanistic perspective in a short period of time. Therefore, it is crucial to provide thermal runaway warnings and take safety measures in advance.

[0003] Currently, there are multiple early warning methods for lithium-ion battery thermal runaway, including monitoring voltage, temperature, gas production, and other signals to predict the occurrence of thermal runaway. However, these signals are prone to misjudgment, susceptibility to interference, delayed alarms, and high monitoring costs. Therefore, it is necessary to further develop an accurate and timely early warning method to minimize the occurrence of thermal runaway. Summary of the Invention

[0004] In view of this, an object of the present invention is to provide a lithium-ion battery thermal runaway warning method based on sound signals.

[0005] In order to achieve the above object, the present invention provides the following technical solutions:

[0006] A lithium-ion battery thermal runaway early warning method based on sound signals comprises the following steps:

[0007] S1: Collect the lithium-ion battery thermal runaway valve opening sound dataset and the environmental noise dataset to establish a battery sound database;

[0008] S2: Use the time series feature extraction library TSFEL to extract the time domain, frequency domain and statistical features of the sound data;

[0009] S3: Grey relational gradient (GRG) is used to analyze the correlation of the proposed features and eliminate redundant features with high correlation;

[0010] S4: Based on the sample data after feature screening, the corresponding sound type is used as a training label to train a random forest classifier based on the Bayesian optimization algorithm to build a battery thermal runaway warning model based on sound signals;

[0011] S5: The test data after feature extraction and selection is used as the input of the battery thermal runaway warning model, and whether the battery has a thermal runaway risk is determined based on whether the valve opening sound is detected.

[0012] Furthermore, step S1 specifically includes: using a sound sensor to collect sound signals generated during battery thermal runaway; collecting lithium-ion battery thermal runaway valve opening sound data and environmental noise data, and establishing a battery sound database.

[0013] Furthermore, step S2 specifically includes the following steps:

[0014] S21: Setting the frame length and frame shift time, and performing frame processing on the collected sound signal;

[0015] S22: Based on the time series feature extraction library TSFEL, configure time domain, frequency domain and statistical feature extraction parameters to perform multi-dimensional feature extraction on the framed sound signal.

[0016] Furthermore, the time domain, frequency domain and statistical feature extraction parameters include: kurtosis, maximum value, skewness, root mean square, zero crossing rate, mean absolute difference, spectral entropy, fundamental frequency, spectral centroid and spectral slope.

[0017] Furthermore, step S3 specifically includes the following steps:

[0018] S31: Grey relational gradient analysis (GRG) is used to perform correlation analysis on the extracted time domain, frequency domain and statistical features, and the correlation between each feature is calculated to screen key acoustic features;

[0019] S32: Based on the correlation analysis results, a correlation threshold is set to filter out redundant features with high correlation, and determine the feature subset that characterizes the key acoustic features of thermal runaway.

[0020] Furthermore, the grey relational gradient analysis method GRG in step S31 combines grey relational analysis with gradient change information to measure the degree of correlation between features. The calculation formula is as follows:

[0021]

[0022] Where, ξ i(k) is the grey relational coefficient between the i-th feature and the other features at the k-th moment, and ξ is the discrimination coefficient.

[0023] Furthermore, in step S32, based on the correlation analysis results, the spectrum centroid and spectrum slope are eliminated, and the kurtosis, maximum value, skewness, root mean square, zero crossing rate, mean absolute difference, spectrum entropy and fundamental frequency are selected as the final feature parameters.

[0024] Further, the step S4 is specifically as follows:

[0025] S41: Label the sample according to its sound type. The noise sample is labeled 0, indicating no thermal runaway state, and the valve opening sound sample is labeled 1, indicating a characteristic acoustic event triggered by battery thermal runaway.

[0026] S42: Divide the feature dataset after feature screening into a training set and a test set, train a random forest classifier based on the Bayesian optimization algorithm, and build a battery thermal runaway warning model based on sound signals.

[0027] Furthermore, the random forest classifier based on the Bayesian optimization algorithm in step S42 first generates an initial set of candidate points and iteratively searches for and adds the next potential extreme point based on the evaluation information of these points. The optimal hyperparameter x best The mathematical expression is:

[0028]

[0029] Where f(x) is the Gaussian distribution function that takes both mean and variance into account, α is the set hyperparameter range, and argmax represents the maximum value of the objective function.

[0030] Further, the step S5 is specifically as follows:

[0031] S51: Inputting the test data after feature extraction and selection into the trained battery thermal runaway warning model to perform valve opening sound recognition;

[0032] S52: Determine whether the battery has a thermal runaway risk by identifying whether there is a valve opening sound in the test data. If there is a thermal runaway risk, record the valve opening time and issue an alert.

[0033] The beneficial effects of the present invention are:

[0034] The thermal runaway warning method based on sound signals monitors the characteristic acoustic signals during battery operation in real time, especially the valve opening acoustic characteristics generated in the early stage of thermal runaway, to accurately judge the events in which lithium-ion batteries have the risk of thermal runaway, and provide early warning for batteries with the risk of thermal runaway.

[0035] 1) The time series feature extraction library (TSFEL) is used to extract multi-dimensional features of acoustic signals. By fusing time domain features, frequency domain features, and statistical features, a multi-dimensional feature vector of sound is constructed to improve the accuracy and robustness of battery thermal runaway valve opening sound recognition.

[0036] 2) The proposed scheme can eliminate the interference of on-site battery operating environment noise, effectively distinguish the environmental background noise from the characteristic acoustic signal of battery thermal runaway, and accurately detect the battery thermal runaway valve opening event, thereby providing timely thermal runaway warning for the lithium-ion battery system and ensuring the safe and stable operation of the battery system.

[0037] 3) The proposed scheme uses a random forest classifier to identify the sound signal of the battery thermal runaway valve opening, and through accurate judgment of the safety valve opening event under battery operating conditions, it provides early warning of battery systems with thermal runaway risks.

[0038] Other advantages, objects, and features of the present invention will be described in part in the following description and, in part, will be apparent to those skilled in the art upon examination of the following description or may be learned from practice of the present invention. The objects and other advantages of the present invention may be realized and obtained through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to make the purpose, technical solutions and advantages of the present invention more clear, the present invention will be described in detail below with reference to the accompanying drawings, in which:

[0040] Figure 1 This is a flow chart of the lithium-ion battery thermal runaway warning method based on sound signals according to the present invention;

[0041] Figure 2 This is a framework diagram of the lithium-ion battery thermal runaway warning algorithm based on sound signals;

[0042] Figure 3 Schematic diagram of feature correlation result analysis;

[0043] Figure 4 This is the test data thermal runaway anomaly detection result. DETAILED DESCRIPTION

[0044] The following describes the embodiments of the present invention by means of specific examples, and those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the illustrations provided in the following embodiments are only schematic illustrations of the basic concept of the present invention, and the following embodiments and features in the embodiments can be combined with each other without conflict.

[0045] It should be noted that the illustrations provided in the following embodiments are merely schematic illustrations of the basic concept of the present invention. Therefore, the illustrations only show components related to the present invention and are not drawn according to the number, shape, and size of components in actual implementation. In actual implementation, the type, quantity, and proportion of each component may be changed arbitrarily, and the component layout may also be more complex.

[0046] In the following description, numerous details are discussed to provide a more thorough explanation of the embodiments of the present invention. However, it will be apparent to those skilled in the art that the embodiments of the present invention may be practiced without these specific details. In other embodiments, well-known structures and devices are shown in block diagram form rather than in detail to avoid obscuring the embodiments of the present invention.

[0047] Example 1:

[0048] like Figure 1 As shown, the present invention provides a lithium-ion battery thermal runaway early warning method based on acoustic signals, which can be divided into the following steps:

[0049] Step S1: Collect a lithium-ion battery thermal runaway valve opening sound dataset and an environmental noise dataset to establish a battery sound database;

[0050] Step S2: Using the Time Series Feature Extraction Library (TSFEL) to extract the time domain, frequency domain and statistical features of the sound data;

[0051] Step S3: Grey relational gradient (GRG) is used to analyze the correlation of the proposed features and eliminate redundant features with high correlation;

[0052] Step S4: Based on the sample data after feature screening, the corresponding sound type is used as a training label to train a random forest classifier based on the Bayesian optimization algorithm to build a battery thermal runaway warning model based on sound signals;

[0053] Step S5: The test data after feature extraction and selection is used as the input of the battery thermal runaway warning model, and whether the battery has a thermal runaway risk is determined based on whether the valve opening sound is detected.

[0054] As an optional embodiment, the complete algorithm framework diagram of this solution is as follows: Figure 2 shown.

[0055] As an optional embodiment, the above step S1 specifically includes steps S11-S12:

[0056] Step S11: Preset a sound sensor to collect sound signals generated during battery thermal runaway at a sampling frequency of 100 kHz, wherein the sound sensor is preferably disposed above the battery;

[0057] Step S12: Collecting lithium-ion battery thermal runaway valve opening sound data and environmental noise data to establish a battery sound database.

[0058] As an optional embodiment, the above step S2 specifically includes steps S21-S22:

[0059] Step S21: Frame the collected sound signal, set the frame length to 50 milliseconds, and the frame shift to 25 milliseconds;

[0060] Step S22: Based on the Time Series Feature Extraction Library (TSFEL), configure time domain, frequency domain and statistical feature extraction parameters to perform multi-dimensional feature extraction on the framed sound signal.

[0061] As an optional embodiment, the time domain, frequency domain and statistical feature extraction parameters include: kurtosis, maximum value, skewness, root mean square, zero crossing rate, mean absolute difference, spectral entropy, fundamental frequency, spectral centroid and spectral slope.

[0062] The features extracted by the present invention include 4 statistical features, 2 time domain features and 4 frequency domain features, and the calculation method is as follows:

[0063] Kurtosis and skewness describe the shape of the sound signal distribution. Kurtosis is used to measure the "sharpness" of the sound signal distribution. Kurtosis indicates that the data is concentrated around the mean and has a thick tail. Skewness is used to measure the degree of asymmetry of the signal distribution:

[0064]

[0065] Where x i is the i-th observation value, μ is the mean, σ is the standard deviation, and n is the number of samples.

[0066] The maximum value indicates the maximum amplitude value that appears in the sound signal and is used to reflect the upper limit of the sound intensity.

[0067] The root mean square is an energy-based feature that measures the overall energy of the signal and is used to evaluate sound intensity:

[0068]

[0069] The zero-crossing rate indicates the number of times a signal crosses zero and is used to distinguish noise from unvoiced sound:

[0070]

[0071] Where, 1 {} It is an indicator function that indicates whether there is a sign change.

[0072] The mean absolute difference is the average amplitude change between consecutive samples, reflecting the stationarity of the signal:

[0073]

[0074] Spectral entropy indicates the randomness or uncertainty of the signal spectrum distribution. A higher spectral entropy indicates a more uniform distribution of sound spectrum energy, such as noise:

[0075]

[0076] Where, P i is the normalized spectral power distribution.

[0077] The fundamental frequency is the lowest frequency of a sound, which usually determines the pitch and is used to detect periodicity and occurrence characteristics.

[0078] The spectrum centroid is the "center of gravity" position of the spectrum, which is used to measure the average position of high-frequency components.

[0079]

[0080] Where, f k is the kth frequency, S k is the corresponding amplitude or power.

[0081] The spectrum slope indicates the decreasing trend of the spectrum with frequency, reflecting the energy attenuation of the signal frequency:

[0082]

[0083] As an optional embodiment, the above step S3 specifically includes steps S31-S32:

[0084] Step S31: using the grey relational gradient analysis (GRG) method to perform correlation analysis on the extracted time domain, frequency domain and statistical features, and calculate the correlation between each feature to screen key acoustic features;

[0085] As an optional embodiment, the grey relational gradient analysis method (GRG) is used to evaluate the correlation between the features and between the features and the battery thermal runaway state, such as Figure 3, which is a schematic diagram of feature correlation result analysis of an embodiment;

[0086] Grey relational gradient analysis (GRG) combines grey relational analysis with gradient change information to measure the degree of correlation between features. The calculation formula is as follows:

[0087]

[0088] Where, ξ i (k) is the grey correlation coefficient between the i-th feature and the other features at the k-th moment, ξ is the discrimination coefficient, which is usually taken as 0.5.

[0089] Step S32: Based on the correlation analysis results, a correlation threshold is set to filter out redundant features with high correlation, and a feature subset representing key acoustic features of thermal runaway is determined.

[0090] As an optional embodiment, in order to eliminate redundant features with high correlation, features with low correlation are selected as inputs of the thermal runaway warning model, and the high correlation threshold is set to 0.9.

[0091] Based on the correlation analysis results, the spectrum centroid and spectrum slope were eliminated, and kurtosis, maximum value, skewness, root mean square, zero-crossing rate, mean absolute difference, spectrum entropy and fundamental frequency were selected as the final feature parameters.

[0092] As an optional embodiment, the above step S4 specifically includes steps S41-S42:

[0093] Step S41: Label the sample according to its type. The noise sample is labeled 0, indicating no thermal runaway state. The valve opening sound sample is labeled 1, indicating a characteristic acoustic event triggered by battery thermal runaway.

[0094] Step S42: Divide the feature data set after feature screening into a training set and a test set, train a random forest classifier, and build a battery thermal runaway warning model based on sound signals.

[0095] As an optional embodiment, 80% of the feature data set is used for the training set, and 20% is used for the test set.

[0096] As an optional embodiment, the random forest classifier integrates the Bayesian optimization algorithm to obtain the optimal value of the hyperparameter, that is, first generate an initial set of candidate points and based on the evaluation information of these points, the algorithm iteratively searches for and adds the next potential extreme point, and the optimal hyperparameter x best The mathematical expression is:

[0097]

[0098] Where f(x) is the Gaussian distribution function that takes both mean and variance into account, α is the set hyperparameter range, and argmax represents the maximum value of the objective function.

[0099] As an optional embodiment, the above step S5 specifically includes steps S51-S52:

[0100] Step S51: Inputting the test data after feature extraction and selection into the trained battery thermal runaway warning model to perform valve opening sound recognition;

[0101] Step S52: Determine whether the battery has a thermal runaway risk by identifying whether there is a valve opening sound in the test data. If there is a thermal runaway risk, record the valve opening time and issue an alert.

[0102] As an optional embodiment, Figure 4 The validation results of the test data in the safety valve opening sound recognition system using the random forest classifier based on the Bayesian optimization algorithm are shown. Figure 4 It can be seen that the system can effectively eliminate environmental noise interference and accurately identify development acoustic characteristics.

[0103] The thermal runaway warning method based on sound signals monitors the characteristic acoustic signals during battery operation in real time, especially the valve opening acoustic characteristics generated in the early stage of thermal runaway, eliminates environmental noise interference, accurately determines events where lithium-ion batteries have the risk of thermal runaway, and issues early warnings for batteries at risk of thermal runaway.

[0104] Example 2:

[0105] An electronic device comprising a memory and a processor;

[0106] The memory is used to store computer programs;

[0107] The processor is configured to implement the method described in Example 1 when executing the computer program.

[0108] Example 3:

[0109] A computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method described in Example 1 is implemented.

[0110] Example 4:

[0111] A computer program product includes a computer program, which implements the method described in embodiment 1 when executed by a processor.

[0112] In the above embodiments, references to "this embodiment" in the specification indicate that a particular feature, structure, or characteristic described in conjunction with the embodiment is included in at least some embodiments, but not necessarily all embodiments. Multiple occurrences of "this embodiment" do not necessarily refer to the same embodiment.

[0113] In the above embodiments, although the invention has been described in conjunction with specific embodiments thereof, many alternatives, modifications, and variations of these embodiments will be apparent to those skilled in the art based on the foregoing description. For example, other memory structures (e.g., dynamic RAM (DRAM)) may be used with the embodiments discussed. The embodiments of the present invention are intended to encompass all such alternatives, modifications, and variations that fall within the broad scope of the appended claims.

[0114] Regarding the computer-readable storage medium in this embodiment, those skilled in the art will appreciate that all or part of the steps in the aforementioned method embodiments can be implemented using hardware associated with the computer program. The aforementioned computer program can be stored in a computer-readable storage medium. When executed, the program performs the steps in the aforementioned method embodiments. The aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks.

[0115] The electronic terminal provided in this embodiment includes a processor, a memory, a transceiver and a communication interface. The memory and the communication interface are connected to the processor and the transceiver and complete communication with each other. The memory is used to store computer programs, the communication interface is used for communication, and the processor and the transceiver are used to run computer programs so that the electronic terminal executes the various steps of the above method.

[0116] In this embodiment, the memory may include a random access memory (RAM), and may also include a non-volatile memory (non-volatile memory), such as at least one disk storage.

[0117] The above-mentioned processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, and discrete hardware components.

[0118] The present invention can be used in a wide variety of general-purpose or special-purpose computing system environments or configurations, such as personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, and distributed computing environments that include any of the above.

[0119] The present invention may be described in the general context of computer-executable instructions, such as program modules, executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, and the like that perform specific tasks or implement specific abstract data types. The present invention may also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via a communications network. In a distributed computing environment, program modules may be located in both local and remote computer storage media, including storage devices.

[0120] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not limiting. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention can be modified or replaced by equivalents without departing from the purpose and scope of the technical solutions, which should all be included in the scope of the claims of the present invention.

Claims

1. A lithium-ion battery thermal runaway early warning method based on sound signals, characterized by: The following steps are involved: S1: Collect the lithium-ion battery thermal runaway valve opening sound dataset and the environmental noise dataset to establish a battery sound database; S2: Use the time series feature extraction library TSFEL to extract the time domain, frequency domain and statistical features of the sound data; S3: Grey relational gradient (GRG) is used to analyze the correlation of the proposed features and eliminate redundant features with high correlation; S4: Based on the sample data after feature screening, the corresponding sound type is used as a training label to train a random forest classifier based on the Bayesian optimization algorithm to build a battery thermal runaway warning model based on sound signals; S5: The test data after feature extraction and selection is used as the input of the battery thermal runaway warning model, and whether the battery has a thermal runaway risk is determined based on whether the valve opening sound is detected.

2. The lithium-ion battery thermal runaway early warning method based on sound signals according to claim 1, characterized in that: Step S1 specifically includes: using a sound sensor to collect sound signals generated during battery thermal runaway; collecting lithium-ion battery thermal runaway valve opening sound data and environmental noise data to establish a battery sound database.

3. The lithium-ion battery thermal runaway early warning method based on sound signals according to claim 1, characterized in that: Step S2 specifically includes the following steps: S21: Setting the frame length and frame shift time, and performing frame processing on the collected sound signal; S22: Based on the time series feature extraction library TSFEL, configure time domain, frequency domain and statistical feature extraction parameters to perform multi-dimensional feature extraction on the framed sound signal.

4. The lithium-ion battery thermal runaway early warning method based on sound signals according to claim 3, characterized in that: The time domain, frequency domain and statistical feature extraction parameters include: kurtosis, maximum value, skewness, root mean square, zero crossing rate, mean absolute difference, spectrum entropy, fundamental frequency, spectrum centroid and spectrum slope.

5. The lithium-ion battery thermal runaway early warning method based on sound signals according to claim 1, characterized in that: Step S3 specifically includes the following steps: S31: Grey relational gradient analysis (GRG) is used to perform correlation analysis on the extracted time domain, frequency domain and statistical features, and the correlation between each feature is calculated to screen key acoustic features; S32: Based on the correlation analysis results, a correlation threshold is set to filter out redundant features with high correlation, and determine the feature subset that characterizes the key acoustic features of thermal runaway.

6. The lithium-ion battery thermal runaway early warning method based on sound signals according to claim 5, characterized in that: The grey relational gradient analysis method GRG in step S31 combines grey relational analysis with gradient change information to measure the degree of correlation between features. The calculation formula is as follows: Where, ξ i (k) is the grey relational coefficient between the i-th feature and the other features at the k-th moment, and ξ is the discrimination coefficient.

7. The lithium-ion battery thermal runaway early warning method based on sound signals according to claim 5, characterized in that: In step S32, based on the correlation analysis results, the spectrum centroid and spectrum slope are eliminated, and the kurtosis, maximum value, skewness, root mean square, zero crossing rate, mean absolute difference, spectrum entropy and fundamental frequency are selected as the final feature parameters.

8. The lithium-ion battery thermal runaway early warning method based on sound signals according to claim 1, characterized in that: The step S4 is specifically as follows: S41: Label the sample according to its sound type. The noise sample is labeled 0, indicating no thermal runaway state, and the valve opening sound sample is labeled 1, indicating a characteristic acoustic event triggered by battery thermal runaway. S42: Divide the feature dataset after feature screening into a training set and a test set, train a random forest classifier based on the Bayesian optimization algorithm, and build a battery thermal runaway warning model based on sound signals.

9. The lithium-ion battery thermal runaway early warning method based on sound signals according to claim 1, characterized in that: The random forest classifier based on the Bayesian optimization algorithm in step S42 first generates an initial set of candidate points and iteratively searches for and adds the next potential extreme point based on the evaluation information of these points. The optimal hyperparameter x best The mathematical expression is: Where f(x) is the Gaussian distribution function that takes both mean and variance into account, α is the set hyperparameter range, and argmax represents the maximum value of the objective function.

10. The lithium-ion battery thermal runaway early warning method based on sound signals according to claim 1, characterized in that: The step S5 is specifically as follows: S51: Inputting the test data after feature extraction and selection into the trained battery thermal runaway warning model to perform valve opening sound recognition; S52: Determine whether the battery has a thermal runaway risk by identifying whether there is a valve opening sound in the test data. If there is a thermal runaway risk, record the valve opening time and issue an alert.

Images