Battery pack thermal runaway risk identification and early warning system and method

Abstract

The invention discloses a battery pack thermal runaway risk identification and early warning system and method, and relates to the related field of battery risk early warning, and the method comprises the steps: collecting an operation parameter data set of a battery pack in real time through a multi-mode sensor network, carrying out the hybrid anomaly detection of the battery pack, obtaining a real-time anomaly detection result, and carrying out the thermal state prediction of the battery pack, and constructing a thermal state evolution trend chart to perform thermal runaway risk assessment, determining a thermal runaway risk level, triggering a multi-level early warning signal according to the thermal runaway risk level, and generating a thermal runaway suppression strategy. The technical problems that in existing battery thermal runaway early warning, due to monitoring limitation of a single sensor, the false alarm rate is high, response lags behind, and battery aging cannot be dynamically adapted are solved, and the technical effects that the thermal runaway early symptom detection rate is increased through multi-mode sensing, and the early warning time is shortened are achieved.

Description

Technical Field

[0001] The present application relates to the field of battery risk warning, and in particular to a battery pack thermal runaway risk identification and warning system and method. Background Art

[0002] With the increase in the energy density of power battery packs, thermal runaway has become a core threat to the safety of lithium-ion batteries. Existing technologies mostly rely on a single sensor (such as temperature or voltage) for threshold judgment, but battery thermal runaway has multi-physical field coupling characteristics (such as gas production, deformation, and local temperature rise), and traditional methods are difficult to capture early signs. In addition, existing early warning systems generally have a response lag problem. The average delay from abnormal detection to triggering protection exceeds 500ms, which cannot meet the real-time requirements of high-rate charging and discharging scenarios.

[0003] The existing technology has technical problems in battery thermal runaway warning, such as high false alarm rate, delayed response and inability to dynamically adapt to battery aging due to the monitoring limitations of a single sensor. Summary of the invention

[0004] The present application provides a battery pack thermal runaway risk identification and warning system and method to solve the technical problems of high false alarm rate, delayed response and inability to dynamically adapt to battery aging caused by the monitoring limitations of a single sensor in the existing battery thermal runaway warning, and achieves the technical effect of improving the detection rate of early signs of thermal runaway and shortening the warning time through multimodal sensing.

[0005] The present application provides a battery pack thermal runaway risk identification and warning system, which is applied to a battery pack thermal runaway risk identification and warning method, including: an acquisition module, which is used to acquire a battery pack operating parameter data set in real time through a multimodal sensor network; a detection module, which is used to perform mixed anomaly detection on the battery pack based on the operating parameter data set to obtain real-time anomaly detection results; a prediction module, which is used to predict the thermal state of the battery pack according to the real-time anomaly detection results and construct a thermal state evolution trend graph; an evaluation module, which is used to perform thermal runaway risk evaluation based on the thermal state evolution trend graph, determine the thermal runaway risk level, trigger a multi-level warning signal according to the thermal runaway risk level, and generate a thermal runaway suppression strategy.

[0006] In a possible implementation, the acquisition module performs the following processing: a first sensing unit is used to distribute an infrared temperature sensor array on the surface of the battery pack and in the gaps between cells to form a first sensing layer; a second sensing unit is used to embed a micro voltage sensor at the connection between the positive and negative electrodes of the battery pack to form a second sensing layer; a third sensing unit is used to install a gas composition spectrometer in the exhaust channel of the battery pack to form a third sensing layer; a fourth sensing unit is used to attach a flexible piezoelectric film sensor to the inner side of the battery pack shell to form a fourth sensing layer; an alignment unit is used to synchronize the first sensing layer, the second sensing layer, the third sensing layer, and the fourth sensing layer according to the operation timestamp to generate the operation parameter data set.

[0007] In a possible implementation, the detection module performs the following processing: a feature extraction unit is used to perform spatial feature extraction based on the operating parameter data set to generate a high-dimensional feature vector; a classification unit is used to perform multi-classification according to the high-dimensional feature vector, build a multi-classification decision, and perform local feature enhancement on the operating parameter data set according to the multi-classification decision to obtain multiple enhanced feature data; a weighted fusion unit is used to retrieve historical detection results, weightedly fuse the multiple enhanced feature data with the historical detection results, and generate a real-time anomaly detection result, wherein the real-time anomaly detection result includes an anomaly type label and a confidence score.

[0008] In a possible implementation, the prediction module performs the following processing: a selection unit is used to dynamically select an input feature subset based on the abnormality detection result to construct a multi-dimensional time series input matrix of the target time window; a thermal runaway prediction unit is used to perform thermal runaway prediction on the multi-dimensional time series input matrix to obtain a thermal state prediction data set, wherein the thermal state prediction data set includes a temperature change rate, a gas production rate, and a deformation acceleration prediction value; a first state evolution unit is used to perform state evolution based on the temperature change rate of the battery pack to obtain spatial temperature distribution data; a second state evolution unit is used to perform state evolution based on the gas production rate of the battery pack to obtain gas diffusion path data; a third state evolution unit is used to perform state evolution based on the deformation acceleration prediction value of the battery pack to obtain mechanical deformation vector data; an association and integration unit is used to associate and integrate the spatial temperature distribution data, the gas diffusion path data, and the mechanical deformation vector data to draw the thermal state evolution trend diagram.

[0009] In a possible implementation, the prediction module performs the following processing: a data screening unit is used to screen the operating parameter data set according to the abnormal type label to generate a first feature subset; a weighting unit is used to weight the first feature subset based on the confidence score to generate a second feature subset; a data extraction unit is used to extract the sliding mean, standard deviation, and kurtosis coefficient of the second feature subset within the target time window to construct a multidimensional statistical feature matrix; a fusion unit is used to perform feature-level fusion of the multidimensional statistical feature matrix with the electrochemical mechanism simulation data to generate the multidimensional time series input matrix.

[0010] In a possible implementation, the evaluation module performs the following processing: a multidimensional decomposition unit is used to perform multidimensional decomposition on the thermal state evolution trend diagram to obtain multidimensional data, wherein the multidimensional data includes heat accumulation rate data, heat diffusion gradient data, and thermal runaway trigger probability data; a dynamic adjustment unit is used to dynamically adjust the weight coefficients of the heat accumulation rate data, the heat diffusion gradient data, and the thermal runaway trigger probability data according to the health status parameters of the battery pack to generate a comprehensive risk score; a risk determination unit is used to determine the risk level according to the comprehensive risk score to obtain the thermal runaway risk level.

[0011] In a possible implementation, the evaluation module performs the following processing: a traversal unit is used to construct a risk level quantization matrix based on the comprehensive risk score, and traverse the risk level quantization matrix to extract the thermal diffusion factor, electrochemical instability index, and deformation energy accumulation value of the battery pack; a threshold setting unit is used to set a multi-level risk threshold, and the multi-level risk threshold includes a first risk threshold, a second risk threshold, and a third risk threshold; a first determination unit is used to determine that it is a level I risk when the thermal diffusion factor exceeds the first risk threshold and the electrochemical instability index is lower than the second risk threshold; a second determination unit is used to determine that it is a level II risk when the thermal diffusion factor exceeds the first risk threshold, the electrochemical instability index exceeds the second risk threshold, and the deformation energy accumulation value is lower than the third risk threshold; a third determination unit is used to determine that it is a level III risk when the thermal diffusion factor exceeds the first risk threshold, the electrochemical instability index exceeds the second risk threshold, and the deformation energy accumulation value exceeds the third risk threshold.

[0012] In a possible implementation, the evaluation module performs the following processing: a dimensional analysis unit is used to analyze the matrix dimension of the risk level quantization matrix to determine the thermodynamic dimension, electrochemical dimension and mechanical dimension; a first calculation unit is used to calculate the product of the maximum temperature difference between adjacent cells in the temperature field and the heat conduction rate as the heat diffusion factor based on the thermodynamic dimension and the heat diffusion gradient data; a second calculation unit is used to obtain the electrochemical instability index based on the electrochemical dimension and the voltage drop rate calculation based on the thermal runaway trigger probability data combined with the gas concentration change curve; a third calculation unit is used to analyze the stress concentration area of ​​the battery pack shell based on the mechanical dimension through the deformation vector field combined with the heat accumulation rate data to calculate the deformation energy accumulation value.

[0013] In a possible implementation, the evaluation module performs the following processing: a fourth calculation unit is used to calculate the three-dimensional risk diffusion of the battery pack through the heat conduction equation according to the real-time temperature distribution data, and determine the thermal runaway propagation path; a first trigger unit is used to trigger the first warning signal when the thermal runaway risk level is level I risk, locate the abnormal position coordinates according to the thermal runaway propagation path, and generate a warning message; a second trigger unit is used to trigger the second warning signal when the thermal runaway risk level is level II risk, dynamically adjust the liquid cooling system according to the thermal runaway propagation path, generate a liquid cooling control instruction, and distribute the flow to the cooling pipeline of the liquid cooling system; a third trigger unit is used to trigger the third warning signal when the thermal runaway risk level is level III risk, activate the directional fire extinguishing device based on the thermal runaway propagation path, and generate an emergency pressure relief channel.

[0014] The present application also provides a method for identifying and warning of thermal runaway risks of battery packs, comprising: collecting an operating parameter data set of the battery pack in real time through a multimodal sensor network; performing mixed anomaly detection on the battery pack based on the operating parameter data set to obtain real-time anomaly detection results; predicting the thermal state of the battery pack based on the real-time anomaly detection results, and constructing a thermal state evolution trend graph; performing a thermal runaway risk assessment based on the thermal state evolution trend graph to determine the thermal runaway risk level, triggering a multi-level warning signal according to the thermal runaway risk level, and generating a thermal runaway suppression strategy.

[0015] One or more technical solutions provided in this application have at least the following technical effects or advantages: The battery pack thermal runaway risk identification and warning system and method provided in the present application relate to the field of battery risk warning technology, and solve the technical problems of high false alarm rate, delayed response and inability to dynamically adapt to battery aging caused by the monitoring limitations of a single sensor in the existing battery thermal runaway warning, and achieve the technical effect of improving the detection rate of early signs of thermal runaway and shortening the warning time through multimodal sensing. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical scheme of the embodiment of the present application, the accompanying drawings of the embodiment of the present application will be briefly introduced below. A flow chart is used in the present application to illustrate the operations performed by the system according to the embodiment of the present application. It should be understood that the preceding or following operations are not necessarily performed accurately in order. On the contrary, various steps can be processed in reverse order or simultaneously as needed. At the same time, other operations can also be added to these processes, or a certain step or several steps of operations can be removed from these processes.

[0017] Figure 1 A schematic diagram of the structure of a battery pack thermal runaway risk identification and warning system provided in an embodiment of the present application; Figure 2 A schematic diagram of a flow chart of a method for identifying and warning of thermal runaway risks in a battery pack provided in an embodiment of the present application. DETAILED DESCRIPTION

[0018] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below.

[0019] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings. The described embodiments should not be regarded as limiting the present application. All other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of this application.

[0020] In the following description, reference is made to "some embodiments", which describe a subset of all possible embodiments, but it is understood that "some embodiments" may be the same subset or different subsets of all possible embodiments, and may be combined with each other without conflict, and the terms "first\second" involved are merely to distinguish similar objects and do not represent a specific ordering of objects. The terms "including" and "having" and any variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or modules that are not clearly listed or inherent to these processes, methods, products, or devices. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those generally understood by technicians in the technical field of this application. The terms used herein are for the purpose of describing the embodiments of the present application only.

[0021] According to the battery pack thermal runaway risk identification and warning system of the embodiment of the present application, Figure 1As shown, it is used to solve the technical problems of high false alarm rate, delayed response and inability to dynamically adapt to battery aging caused by the monitoring limitations of a single sensor in the existing battery thermal runaway warning, and to achieve the technical effect of improving the detection rate of early signs of thermal runaway through multi-modal sensing and shortening the warning time. The battery pack thermal runaway risk identification and warning system includes: an acquisition module 10, a detection module 20, a prediction module 30, and an evaluation module 40.

[0022] The acquisition module 10 is used to collect the operating parameter data set of the battery pack in real time through a multimodal sensor network; the specific configuration of the acquisition module 10 will be described in detail below. As described above, the acquisition module 10 may further include: a first sensing unit, used to distribute the infrared temperature sensor array on the surface of the battery pack and the gap between the cells to form a first sensing layer; a second sensing unit, used to embed a micro voltage sensor at the connection between the positive and negative electrodes of the battery pack to form a second sensing layer; a third sensing unit, used to install a gas composition spectrometer in the exhaust channel of the battery pack to form a third sensing layer; a fourth sensing unit, used to attach a flexible piezoelectric film sensor to the inner side of the battery pack shell to form a fourth sensing layer; an alignment unit, used to synchronize the first sensing layer, the second sensing layer, the third sensing layer, and the fourth sensing layer according to the operating timestamp to generate the operating parameter data set.

[0023] First, for the first sensing layer, a high-density infrared temperature sensor array (such as FLIR A315, accuracy ±1°C) is used on the surface of the battery module and the gap between the cells, arranged in a matrix with a spacing of 5mm, and each single cell corresponds to 3 temperature measurement points. The signal transmission line is integrated through a flexible FPC circuit board, and the sampling frequency of the sensor array is set to 10Hz, and it is fixed to the battery surface by thermal conductive glue to reduce the contact thermal resistance. The micro voltage sensor of the second sensing layer (such as TI INA226, resolution 0.1mV) adopts an embedded installation process, slots are cut at the positive and negative poles to embed the sensor probe, and the surface is covered with an insulating ceramic coating (thickness ≤0.2mm). Differential measurement technology is used to eliminate the influence of contact resistance, and the sampling frequency is set to 10Hz synchronously with the temperature sensing layer. The gas analyzer of the third sensing layer (such as Ocean Insight HR2000+ spectrometer) is installed in the main exhaust channel of the battery pack, and in-situ detection is performed through a quartz glass window. A gas pump is equipped to establish a gas circulation path at a flow rate of 200mL / min, and the characteristic absorption peaks of characteristic gases such as CO and HF are monitored in real time. The spectral scanning interval is set to 500ms. The flexible piezoelectric film sensor of the fourth sensing layer (such as Measurement Specialties LDT0-028K) is made of a special-shaped structure that matches the curved surface of the battery shell by laser cutting technology, and is attached to the inner side of the shell by conductive silver glue. The sampling rate is set to 100Hz to capture micro-deformation signals (resolution 0.1μm). The time synchronization module uses the AD9548 clock generator chip to provide a unified clock source for each sensing layer through the IEEE 1588 Precision Time Protocol (PTP). A timestamp marking mechanism is established in the FPGA processing module (Xilinx Zynq UltraScale+), and the CAN FD bus is used to transmit data frames and attach 8-byte timestamps (accuracy ±1μs). Finally, the time synchronization module is used to align the data of each sensing layer at the millisecond level, which means that the four-layer sensor data is interpolated and resampled through the time alignment algorithm to generate a synchronized operating parameter data set with a time deviation of less than 2ms. The data storage format adopts a hierarchical HDF5 structure, which contains four groups of two-dimensional arrays with timestamps (time × space coordinates) of temperature, voltage, gas concentration, and deformation to determine the operating parameter data set.

[0024] The detection module 20 is used to perform mixed anomaly detection on the battery pack based on the operating parameter data set to obtain real-time anomaly detection results; the specific configuration of the detection module 20 will be described in detail below. As described above, the detection module 20 may further include: a feature extraction unit, used to perform spatial feature extraction based on the operating parameter data set to generate a high-dimensional feature vector; a classification unit, used to perform multi-classification according to the high-dimensional feature vector, build a multi-classification decision, perform local feature enhancement on the operating parameter data set according to the multi-classification decision, and obtain multiple enhanced feature data; a weighted fusion unit, used to retrieve historical detection results, perform weighted fusion on the multiple enhanced feature data and the historical detection results, and generate real-time anomaly detection results, wherein the real-time anomaly detection results include anomaly type labels and confidence scores.

[0025] The spatial feature extraction and anomaly detection are performed by using a three-dimensional convolutional neural network (3D-CNN). First, the four-layer sensor data is constructed into a four-dimensional input tensor (time × spatial coordinate × sensor layer × physical quantity). The cross-sensor layer feature fusion is performed through three layers of 3×3×3 convolution kernels. The first layer of convolution extracts the local temperature-voltage correlation feature (the activation function is ReLU), the second layer of convolution captures the gas-deformation coupling feature (the SELU activation function is used to enhance the nonlinear expression ability), and the third layer of convolution generates a 128-dimensional high-dimensional feature vector. Then, the feature vector is input into the improved support vector machine (SVM) classifier, and the Mahalanobis distance weighted multi-classification decision tree is used to classify the anomaly. The common types are divided into three categories: normal, early thermal runaway (gas production-dominated, temperature rise-dominated), and severe anomalies. A sliding window mechanism (window length 5 seconds, step length 1 second) is introduced in the classification process. The feature vectors of continuous time windows are residually connected, and the temporal feature enhancement is achieved through the gated recurrent unit (GRU). Enhanced feature data containing spatiotemporal correlation is generated, that is, real-time anomaly detection results, and historical detection results are stored in the ring buffer database of the edge computing unit (capacity retains the latest 72 hours of data). The exponential decay weighted algorithm (attenuation coefficient λ=0.95) is used to fuse the current enhanced features with the historical data. Specifically, the joint confidence is calculated through the Bayesian probability framework: ; in, For the current enhanced features, For historical features, the final anomaly type label is determined by the maximum posterior probability, and the confidence score is normalized and output by the Softmax function.

[0026] The prediction module 30 is used to predict the thermal state of the battery pack according to the real-time abnormality detection result and construct a thermal state evolution trend diagram; the specific configuration of the prediction module 30 will be described in detail below. As described above, the prediction module 30 may further include: a selection unit, which is used to dynamically select an input feature subset based on the abnormality detection result to construct a multi-dimensional time series input matrix of the target time window; a thermal runaway prediction unit, which is used to perform thermal runaway prediction on the multi-dimensional time series input matrix to obtain a thermal state prediction data set, and the thermal state prediction data set includes a temperature change rate, a gas production rate, and a deformation acceleration prediction value; a first state evolution unit, which is used to perform state evolution based on the temperature change rate of the battery pack to obtain spatial temperature distribution data; a second state evolution unit, which is used to perform state evolution based on the gas production rate of the battery pack to obtain gas diffusion path data; a third state evolution unit, which is used to perform state evolution based on the deformation acceleration prediction value of the battery pack to obtain mechanical deformation vector data; an association integration unit, which is used to associate and integrate the spatial temperature distribution data, the gas diffusion path data, and the mechanical deformation vector data to draw the thermal state evolution trend diagram.

[0027] Based on the anomaly type labels (such as gas production-dominated or temperature rise-dominated) and confidence scores in the real-time anomaly detection results, the Mahalanobis distance and mutual information method are used to dynamically screen the key feature subsets. For example, when gas production anomalies are detected (confidence>0.8), the three types of features, gas concentration, temperature gradient and voltage fluctuation rate, are prioritized. A multi-dimensional time series input matrix (dimension: time step × number of features × spatial nodes) is constructed through a sliding time window (length 5 minutes, step length 10 seconds), and the theoretical heat production rate output by the electrochemical model is introduced as a physical constraint feature to achieve feature-level fusion.

[0028] The input matrix is ​​further input into the improved Bi-LSTM prediction network, and the network structure includes a bidirectional LSTM layer (128 nodes per layer) and an attention mechanism layer. The temperature prediction branch uses a three-dimensional finite element heat conduction model for physical constraints, and predicts the temperature field distribution for the next 2 minutes every 10 seconds (spatial resolution 1cm³); the gas production prediction branch combines CFD simulation data, calculates the gas diffusion path through the concentration gradient, and outputs the main diffusion direction and velocity vector; the deformation prediction branch uses finite element analysis to calculate the shell stress distribution and outputs the deformation acceleration vector field.

[0029] The three types of prediction data are mapped to a unified coordinate system through a time-space alignment algorithm. The temperature field is quickly interpolated using an octree spatial index, the gas path is generated into streamline trajectories using the particle tracking method, and the deformation vector is transformed using a Lagrangian description. The visualization engine (based on the PyVista library) constructs a three-dimensional dynamic trend chart, the process of which may include mapping the predicted temperature data to the three-dimensional grid model of the battery pack, using the HSV color space (red→blue corresponds to 45℃→80℃) to represent the temperature distribution, superimposing isothermal surfaces (intervals of 5℃) to display the heat accumulation area, and then driving the SmoothedParticle Hydrodynamics (SPH) fluid dynamics model according to the predicted value of the gas production rate to generate a gas cloud with a transparency gradient (CO is displayed in orange and HF is displayed in purple). The length of the streamline arrow is proportional to the diffusion rate. The deformation vector arrow is drawn on the surface of the battery shell (scaling factor 50:1), the direction of the arrow indicates the deformation trend, and the color depth (green→red) reflects the magnitude of the deformation acceleration (0→5mm / s²). The Marching Cubes algorithm is used to extract the intersection area with temperature>60℃, gas production rate>50ppm / s and deformation acceleration>3mm / s², and the core area of ​​thermal runaway is highlighted with pulsed red light. The system updates the trend chart every 30 seconds and supports user interactive operations (rotation / zoom / section view). At the same time, key parameters (such as maximum temperature rise rate and total gas production) are superimposed and displayed in the form of digital instruments. When the deviation between the predicted data and the measured value exceeds 15%, the online fine-tuning of the LSTM network (learning rate 0.001, batch size 32) is automatically triggered to obtain the thermal state evolution trend chart.

[0030] The specific configuration of the prediction module 30 will be described in detail below. As described above, the prediction module 30 may further include: a data screening unit, for screening from the operating parameter data set according to the abnormal type label to generate a first feature subset; a weighting unit, for weighting the first feature subset based on the confidence score to generate a second feature subset; a data extraction unit, for extracting the sliding mean, standard deviation, and kurtosis coefficient of the second feature subset within the target time window to construct a multidimensional statistical feature matrix; a fusion unit, for performing feature-level fusion of the multidimensional statistical feature matrix with the electrochemical mechanism simulation data to generate the multidimensional time series input matrix.

[0031] From the operating parameter data set, we screened according to the anomaly type label and extracted the parameters closely related to the specific anomaly, that is, the sensor data whose correlation exceeded the set threshold, to form the first feature subset. At the same time, we introduced a confidence scoring mechanism to weight each feature in the first feature subset. The confidence score reflects the reliability of the feature in predicting the risk of thermal runaway. By weighting the influence of different features in risk assessment, we generated the second feature subset.

[0032] Subsequently, within the target time window, the second feature subset is statistically analyzed to calculate statistical indicators such as the sliding mean, standard deviation, and kurtosis coefficient. The sliding mean is obtained by setting a fixed-size window on the data sequence and calculating the arithmetic mean of the data in the window. As the window slides on the data sequence, a series of average values ​​can be obtained, thereby reflecting the local trend of the data.

[0033] The standard deviation is a statistic that measures the degree of dispersion of data. It is obtained by averaging the squares of the differences between each value in the data sequence and the mean, and then taking the square root. The larger the standard deviation, the more dispersed the data points are, and vice versa, the more concentrated the data points are.

[0034] The kurtosis coefficient is a statistic that describes the distribution of data. It reflects the sharpness of the data distribution. The calculation of the kurtosis coefficient involves the average of the fourth power of the difference between each value in the data sequence and the mean value, divided by the fourth power of the standard deviation. When the kurtosis coefficient is greater than 3, it means that the data distribution is sharper than the normal distribution; when it is less than 3, it means that the data distribution is flatter. The calculation of these statistics provides an important basis for data analysis.

[0035] The above indicators reveal the changing patterns and potential risks of the characteristics from the perspectives of stability, discreteness and distribution morphology of the time series. On this basis, a multidimensional statistical feature matrix is ​​constructed. In order to further improve the accuracy of risk assessment, the multidimensional statistical feature matrix is ​​fused with the electrochemical mechanism simulation data at the feature level. This process integrates the two at the feature level. First, the multidimensional statistical feature matrix and the electrochemical mechanism simulation data are standardized to ensure that they are numerically comparable. Then, through specific fusion algorithms, such as splicing, weighted averaging or more complex machine learning algorithms, the two are fused in the feature space, which not only combines the real-time and accuracy of the experimental data, but also incorporates the in-depth mechanism understanding of the simulation data, thereby generating a multidimensional time series input matrix containing rich information and deep mechanisms as the input of the subsequent risk assessment model, providing more reliable protection for the safe operation of the battery pack.

[0036] The evaluation module 40 is used to perform a thermal runaway risk evaluation based on the thermal state evolution trend diagram, determine the thermal runaway risk level, trigger a multi-level warning signal according to the thermal runaway risk level, and generate a thermal runaway suppression strategy.

[0037] The specific configuration of the evaluation module 40 will be described in detail below. As described above, the evaluation module 40 may further include: a multidimensional decomposition unit, which is used to multidimensionally decompose the thermal state evolution trend diagram to obtain multidimensional data, wherein the multidimensional data includes heat accumulation rate data, heat diffusion gradient data, and thermal runaway trigger probability data; a dynamic adjustment unit, which is used to dynamically adjust the weight coefficients of the heat accumulation rate data, the heat diffusion gradient data, and the thermal runaway trigger probability data according to the health status parameters of the battery pack to generate a comprehensive risk score; and a risk determination unit, which is used to determine the risk level according to the comprehensive risk score to obtain the thermal runaway risk level.

[0038] Based on the four-dimensional space-time data (three-dimensional space + time) of the thermal state evolution trend diagram, tensor decomposition technology is used to extract key risk parameters, and the local temperature change rate is calculated through space-time derivative operations. The heat accumulation rate data is determined by combining Gaussian filtering to eliminate noise, and then the three-dimensional heat conduction tensor is further constructed. The finite difference method is used to solve the anisotropic diffusion coefficient to determine the thermal diffusion gradient data, and the gas production rate and deformation acceleration are integrated to construct the regression coefficient to determine the thermal runaway trigger probability data.

[0039] An adaptive weight matrix based on the battery state of health (SOH) is further established. When SOH>80%, the heat accumulation rate weight w1=0.5, the diffusion gradient w2=0.3, and the trigger probability w3=0.2. When 40%≤SOH≤80%, the weights are dynamically adjusted according to w1=0.3+0.2×(SOH / 40), w2=0.4, and w3=0.3−0.1×(SOH / 40). When SOH<40%, the trigger probability weight is increased to w3=0.5, and the remaining weights are proportionally reduced to generate a comprehensive risk score. The risk level is determined based on the comprehensive risk score, thereby obtaining the thermal runaway risk level of the battery pack. A comprehensive, accurate and dynamic assessment of the thermal runaway risk of the battery pack is performed to ensure the safe and stable operation of the battery pack.

[0040] The specific configuration of the evaluation module 40 will be described in detail below. As described above, the evaluation module 40 may further include: a traversal unit, which is used to construct a risk level quantization matrix based on the comprehensive risk score, and traverse the risk level quantization matrix to extract the thermal diffusion factor, electrochemical instability index, and deformation energy accumulation value of the battery pack; a threshold setting unit, which is used to set a multi-level risk threshold, and the multi-level risk threshold includes a first risk threshold, a second risk threshold, and a third risk threshold; a first determination unit, which is used to determine that it is a level I risk when the thermal diffusion factor exceeds the first risk threshold and the electrochemical instability index is lower than the second risk threshold; a second determination unit, which is used to determine that it is a level II risk when the thermal diffusion factor exceeds the first risk threshold, the electrochemical instability index exceeds the second risk threshold, and the deformation energy accumulation value is lower than the third risk threshold; a third determination unit, which is used to determine that it is a level III risk when the thermal diffusion factor exceeds the first risk threshold, the electrochemical instability index exceeds the second risk threshold, and the deformation energy accumulation value exceeds the third risk threshold.

[0041] Based on the three-dimensional thermal state evolution data of the battery pack, a three-dimensional quantitative matrix including thermodynamics, electrochemistry, and mechanical mechanics is established. ) is calculated using the improved Green's function method: ; in, is the maximum temperature difference between adjacent cells, is the heat generation rate of the kth heat source, is the heat source distance, is the thermal conductivity attenuation coefficient of the material, is the effective heat dissipation surface area.

[0042] Electrochemical instability index ( ) is calculated by coupling the gas concentration mutation rate with the voltage relaxation time: ; in, is the electrochemical coupling coefficient, is the voltage recovery time constant.

[0043] Deformation energy accumulation value ( ) is integrated through the shell strain energy density: ; in, is the stress tensor, is the strain tensor.

[0044] Furthermore, by adopting the sliding window statistical method, the first risk threshold can be updated every 5 minutes, the second risk threshold can be determined based on the Arrhenius equation correction, and the third risk threshold can be determined by associating the fatigue characteristics of the material. When the thermal diffusion factor exceeds the first risk threshold and the electrochemical instability index is lower than the second risk threshold, it is determined as a level I risk, that is, a local overheating risk, triggering a yellow warning (HMI interface flashing warning); when the thermal diffusion factor exceeds the first risk threshold, the electrochemical instability index exceeds the second risk threshold and the accumulated deformation energy value is lower than the third risk threshold, it is determined as a level II risk, that is, a thermal-electric coupling risk, triggering an orange warning; when the thermal diffusion factor exceeds the first risk threshold, the electrochemical instability index exceeds the second risk threshold and the accumulated deformation energy value exceeds the third risk threshold, it is determined as a level III risk, that is, when all three indicators exceed the threshold, the system triggers a red warning within 10ms.

[0045] The specific configuration of the evaluation module 40 will be described in detail below. As described above, the evaluation module 40 may further include: a dimensional analysis unit for analyzing the matrix dimension of the risk level quantification matrix to determine the thermodynamic dimension, electrochemical dimension and mechanical dimension; a first calculation unit for calculating the product of the maximum temperature difference between adjacent cells in the temperature field and the heat conduction rate as the heat diffusion factor based on the thermodynamic dimension and the heat diffusion gradient data; a second calculation unit for calculating the voltage drop rate based on the electrochemical dimension and the thermal runaway trigger probability data combined with the gas concentration change curve to obtain the electrochemical instability index; a third calculation unit for calculating the deformation energy accumulation value based on the mechanical dimension by analyzing the stress concentration area of ​​the battery pack shell through the deformation vector field combined with the heat accumulation rate data.

[0046] In the construction of the risk level quantification matrix and parameter calculation, the system realizes three-dimensional risk assessment through multi-physical field coupling analysis, including thermodynamic dimension, electrochemical dimension, and mechanical dimension. It further constructs the thermal network model between battery cells based on the three-dimensional temperature field data collected by the infrared temperature sensor array (spatial resolution 1cm³). The calculation process of the thermal diffusion factor is as follows: All adjacent pairs of cells in the three-dimensional temperature field data are traversed, and the real-time temperature difference of all adjacent pairs of cells is calculated. The maximum temperature difference value is recorded, and then the heat conduction rate is obtained by correction through Fourier's law combined with the change of contact thermal resistance caused by aging. Finally, the maximum temperature difference value is multiplied by the heat conduction rate and used as the heat diffusion factor. The higher the heat diffusion factor, the stronger the thermal diffusion capacity of the battery pack.

[0047] Furthermore, based on the electrochemical dimension, the electrochemical instability index is calculated by combining the gas spectrum data (CO / HF concentration CgasCgas) and the voltage sensor data: the gas concentration is monitored in real time by using the CUSUM algorithm, and an event marker is triggered when the rate of change of the gas concentration exceeds the baseline to obtain the gas concentration mutation detection result. At the same time, the voltage change rate is calculated within a 1-second time window, and the relaxation time constant is extracted by Hilbert transform. The exponential synthesis formula of the electrochemical instability index is: ; in, is the electrochemical instability index, is the gas concentration, is the critical value of gas concentration, is the relaxation time constant, is the reference relaxation time under healthy conditions.

[0048] Furthermore, based on the mechanical dimension, the deformation energy accumulation value is constructed through the strain data (sampling rate 1kHz) of the flexible piezoelectric film sensor, which means that based on the shell finite element model, the discrete strain measurement value is interpolated into the full-field strain tensor, and the stress tensor is calculated according to the anisotropic parameters of the composite material. The stress tensor is obtained by multiplying the full-field strain tensor by the stiffness matrix. The heat accumulation rate can also be introduced to compensate the temperature of the Young's modulus, and the mechanical energy integral is determined by GPU acceleration (CUDA kernel function), which can achieve a processing speed of 1M units / ms. All dimensional data are aligned by timestamp (error <1ms), and the quaternion interpolation method is used to compensate for sensor delay. This solution reduces the time consumption of three-dimensional risk assessment from 120ms of the traditional method to 18ms, and at the same time controls the error of cross-dimensional coupling analysis within 3%.

[0049] The specific configuration of the evaluation module 40 will be described in detail below. As described above, the evaluation module 40 may further include: a fourth calculation unit, which is used to calculate the three-dimensional risk diffusion of the battery pack through the heat conduction equation according to the real-time temperature distribution data, and determine the thermal runaway propagation path; a first trigger unit, which is used to trigger the first warning signal when the thermal runaway risk level is level I risk, locate the abnormal position coordinates according to the thermal runaway propagation path, and generate a warning message; a second trigger unit, which is used to trigger the second warning signal when the thermal runaway risk level is level II risk, dynamically adjust the liquid cooling system according to the thermal runaway propagation path, generate a liquid cooling control instruction, and distribute the flow to the cooling pipeline of the liquid cooling system; a third trigger unit, which is used to trigger the third warning signal when the thermal runaway risk level is level III risk, activate the directional fire extinguishing device based on the thermal runaway propagation path, and generate an emergency pressure relief channel.

[0050] Based on real-time infrared temperature data (sampling rate 10Hz), a three-dimensional non-uniform heat conduction model of the battery pack is constructed. The heat diffusion process can be described by using a variable coefficient parabolic PDE, and the thermal conductivity of the battery pack varies with SOH and temperature. The battery pack is then discretized into 5 million tetrahedral units, and the explicit Euler method is used to solve the temperature field evolution in parallel. The three-dimensional temperature distribution is updated every 50ms. At the same time, the Dijkstra algorithm is used to find the path with the maximum temperature gradient, and the thermal runaway propagation direction vector is defined. The paths with a propagation rate exceeding 0.5m / s are marked as high-risk channels, and the thermal runaway propagation path is determined.

[0051] Furthermore, when the thermal runaway risk level is level I risk, the first warning signal is triggered, and the coordinates of the abnormal position are located according to the thermal runaway propagation path, which means that the hotspot coordinates are detected by the extreme value of the second-order derivative of the temperature field, and the positioning accuracy is ±3mm. A JSON format message is constructed to generate a warning message, and the three axial flow fans closest to the hotspot are activated (wind speed v=5+10×(Score−0.4)v=5+10×(Score−0.4) m / s), and the charging current is limited through the BMS.

[0052] Furthermore, when the thermal runaway risk level is level II risk, a second warning signal is triggered, and the liquid cooling system is dynamically adjusted according to the thermal runaway propagation path. This means that based on the propagation path topology, the liquid cooling pipeline is divided into a priority cooling zone (upstream of the path) and a secondary cooling zone (downstream of the path), and liquid cooling control instructions are generated. The flow distribution is performed on the cooling pipeline of the liquid cooling system. The flow distribution can be controlled by using a PID-Smith predictor. When the local temperature rise rate is >3°C / s, it is switched to a turbulent mode (Reynolds number Re>4000), thereby increasing the heat transfer coefficient by 30%.

[0053] Furthermore, when the thermal runaway risk level is level III risk, the third warning signal is triggered, and the directional fire extinguishing device is activated based on the thermal runaway propagation path, which means that the fire extinguishing agent injection angle is calculated according to the propagation path vector, and the angle control accuracy is ±0.5°, the response time is <50ms, and the fire extinguishing agent flow is controlled by pulse width modulation (PWM). Then, the pressure gradient field is reversely solved based on the shell deformation data, and the optimal pressure relief point sequence is determined. The shape memory alloy (SMA) driver is used to open the pressure relief valve. The opening sequence satisfies the optimal pressure relief point sequence to ensure that the pressure release wave front meets the heat propagation wave front in reverse, generating an emergency pressure relief channel. The response action time of the emergency pressure relief channel is compressed from the traditional 150ms to 35ms, and the fire extinguishing agent coverage efficiency is increased to 92%.

[0054] The embodiments of the present application solve the technical problems of high false alarm rate, delayed response and inability to dynamically adapt to battery aging caused by the monitoring limitations of a single sensor in the existing battery thermal runaway warning, and achieve the technical effect of improving the detection rate of early signs of thermal runaway and shortening the warning time through multimodal sensing.

[0055] In the above, refer to Figure 1 The battery pack thermal runaway risk identification and warning system according to an embodiment of the present application is described in detail. Figure 2 A method for identifying and warning of thermal runaway risk of a battery pack according to an embodiment of the present application is described.

[0056] The present application embodiment provides a method for identifying and warning a thermal runaway risk of a battery pack, the method being applied to a thermal runaway risk identification and warning system for a battery pack, the method comprising: The operating parameter data set of the battery pack is collected in real time through a multimodal sensor network; based on the operating parameter data set, mixed anomaly detection is performed on the battery pack to obtain real-time anomaly detection results; based on the real-time anomaly detection results, the thermal state of the battery pack is predicted, and a thermal state evolution trend diagram is constructed; based on the thermal state evolution trend diagram, a thermal runaway risk assessment is performed to determine the thermal runaway risk level, a multi-level warning signal is triggered according to the thermal runaway risk level, and a thermal runaway suppression strategy is generated.

[0057] The deployment process of a multimodal sensor network includes: An infrared temperature sensor array is distributedly deployed on the surface of the battery pack and in the gaps between cells to form a first sensing layer; a micro voltage sensor is embedded at the connection between the positive and negative electrodes of the battery pack to form a second sensing layer; a gas composition spectrometer is installed in the exhaust channel of the battery pack to form a third sensing layer; a flexible piezoelectric film sensor is attached to the inner side of the battery pack shell to form a fourth sensing layer; the first sensing layer, the second sensing layer, the third sensing layer, and the fourth sensing layer are time-synchronized and aligned according to the operation timestamp to generate the operation parameter data set.

[0058] Performing hybrid anomaly detection on the battery pack based on the operating parameter data set to obtain real-time anomaly detection results includes: Based on the operating parameter data set, spatial feature extraction is performed to generate a high-dimensional feature vector; multi-classification is performed based on the high-dimensional feature vector, a multi-classification decision is constructed, and local feature enhancement is performed on the operating parameter data set according to the multi-classification decision to obtain multiple enhanced feature data; historical detection results are retrieved, and the multiple enhanced feature data are weightedly fused with the historical detection results to generate real-time anomaly detection results, which include anomaly type labels and confidence scores.

[0059] Predicting the thermal state of the battery pack according to the real-time abnormality detection result and constructing a thermal state evolution trend diagram includes: Based on the abnormal detection results, a subset of input features is dynamically selected to construct a multi-dimensional time series input matrix of a target time window; thermal runaway is predicted on the multi-dimensional time series input matrix to obtain a thermal state prediction data set, which includes a temperature change rate, a gas production rate, and a deformation acceleration prediction value; state evolution is performed based on the temperature change rate of the battery pack to obtain spatial temperature distribution data; state evolution is performed based on the gas production rate of the battery pack to obtain gas diffusion path data; state evolution is performed based on the deformation acceleration prediction value of the battery pack to obtain mechanical deformation vector data; the spatial temperature distribution data, the gas diffusion path data, and the mechanical deformation vector data are correlated and integrated to draw the thermal state evolution trend diagram.

[0060] Dynamically select an input feature subset based on the anomaly detection result to construct a multi-dimensional time series input matrix of the target time window, including: The operating parameter data set is screened according to the abnormal type label to generate a first feature subset; the first feature subset is weighted based on the confidence score to generate a second feature subset; the sliding mean, standard deviation, and kurtosis coefficient of the second feature subset within a target time window are extracted to construct a multidimensional statistical feature matrix; the multidimensional statistical feature matrix is ​​fused with the electrochemical mechanism simulation data at the feature level to generate the multidimensional time series input matrix.

[0061] Based on the thermal state evolution trend diagram, a thermal runaway risk assessment is performed to determine the thermal runaway risk level, including: The thermal state evolution trend graph is multi-dimensionally decomposed to obtain multi-dimensional data, wherein the multi-dimensional data includes heat accumulation rate data, heat diffusion gradient data, and thermal runaway trigger probability data; weight coefficients of the heat accumulation rate data, the heat diffusion gradient data, and the thermal runaway trigger probability data are dynamically adjusted according to health status parameters of the battery pack to generate a comprehensive risk score; and risk level determination is performed according to the comprehensive risk score to obtain the thermal runaway risk level.

[0062] The risk level is determined according to the comprehensive risk score to obtain the thermal runaway risk level, including: A risk level quantification matrix is ​​constructed based on the comprehensive risk score, and the thermal diffusion factor, electrochemical instability index, and deformation energy accumulation value of the battery pack are extracted by traversing the risk level quantification matrix; a multi-level risk threshold is set, and the multi-level risk threshold includes a first risk threshold, a second risk threshold, and a third risk threshold; when the thermal diffusion factor exceeds the first risk threshold and the electrochemical instability index is lower than the second risk threshold, it is determined to be a level I risk; when the thermal diffusion factor exceeds the first risk threshold, the electrochemical instability index exceeds the second risk threshold, and the deformation energy accumulation value is lower than the third risk threshold, it is determined to be a level II risk; when the thermal diffusion factor exceeds the first risk threshold, the electrochemical instability index exceeds the second risk threshold, and the deformation energy accumulation value exceeds the third risk threshold, it is determined to be a level III risk.

[0063] Traversing the risk level quantification matrix to extract the thermal diffusion factor, electrochemical room temperature index, and deformation energy accumulation value of the battery pack, including: Analyze the matrix dimensions of the risk level quantification matrix to determine the thermodynamic dimension, electrochemical dimension and mechanical dimension; based on the thermodynamic dimension, calculate the product of the maximum temperature difference between adjacent cells in the temperature field and the heat conduction rate according to the thermal diffusion gradient data as the thermal diffusion factor; based on the electrochemical dimension, calculate the electrochemical instability index based on the thermal runaway trigger probability data combined with the gas concentration change curve voltage drop rate; based on the mechanical dimension, analyze the stress concentration area of ​​the battery pack shell through the deformation vector field combined with the heat accumulation rate data, and calculate the deformation energy accumulation value.

[0064] According to the thermal runaway risk level, a multi-level warning signal is triggered, including: The three-dimensional risk diffusion of the battery pack is calculated through the heat conduction equation according to the real-time temperature distribution data to determine the thermal runaway propagation path; when the thermal runaway risk level is level I risk, the first warning signal is triggered, the abnormal position coordinates are located according to the thermal runaway propagation path, and a warning message is generated; when the thermal runaway risk level is level II risk, the second warning signal is triggered, the liquid cooling system is dynamically adjusted according to the thermal runaway propagation path, a liquid cooling control instruction is generated, and the flow is distributed to the cooling pipeline of the liquid cooling system; when the thermal runaway risk level is level III risk, the third warning signal is triggered, the directional fire extinguishing device is activated based on the thermal runaway propagation path, and an emergency pressure relief channel is generated.

[0065] The battery pack thermal runaway risk identification and warning system provided in the embodiments of the present application can execute the battery pack thermal runaway risk identification and warning method provided in any embodiment of the present application, and has the corresponding functional modules and beneficial effects of the execution method.

[0066] Although the present application makes various references to certain modules in the system according to the embodiments of the present application, any number of different modules may be used and run on the user terminal and / or server, and the various units and modules included are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be achieved; in addition, the specific names of the functional units are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of the present application.

[0067] The above specific implementations do not constitute a limitation on the protection scope of this application. It should be understood by those skilled in the art that various modifications, combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of this application should be included in the protection scope of this application.

Claims

1. Battery pack thermal runaway risk identification and warning system, characterized by: The system comprises: A collection module, used for collecting the operating parameter data set of the battery pack in real time through a multimodal sensor network; A detection module, configured to perform mixed anomaly detection on the battery pack based on the operating parameter data set to obtain a real-time anomaly detection result; A prediction module, used to predict the thermal state of the battery pack according to the real-time abnormality detection result and construct a thermal state evolution trend diagram; An evaluation module is used to perform a thermal runaway risk assessment based on the thermal state evolution trend diagram, determine the thermal runaway risk level, trigger a multi-level warning signal according to the thermal runaway risk level, and generate a thermal runaway suppression strategy.

2. The battery pack thermal runaway risk identification and warning system according to claim 1, characterized in that: Acquisition module, including: The first sensing unit is used to distribute an infrared temperature sensor array on the surface of the battery pack and the gaps between cells to form a first sensing layer; A second sensing unit is used to embed a micro voltage sensor at the connection between the positive and negative electrodes of the battery pack to form a second sensing layer; A third sensing unit is used to install a gas composition spectrometer in the exhaust channel of the battery pack to form a third sensing layer; A fourth sensing unit, used for attaching a flexible piezoelectric film sensor to the inner side of the battery pack housing to form a fourth sensing layer; An alignment unit is used to perform time synchronization alignment on the first sensing layer, the second sensing layer, the third sensing layer, and the fourth sensing layer according to an operation timestamp to generate the operation parameter data set.

3. The battery pack thermal runaway risk identification and warning system according to claim 1, characterized in that: Detection module, including: A feature extraction unit, used to extract spatial features based on the operating parameter data set to generate a high-dimensional feature vector; A classification unit, used for performing multi-classification according to the high-dimensional feature vector, constructing a multi-classification decision, and performing local feature enhancement on the operating parameter data set according to the multi-classification decision to obtain a plurality of enhanced feature data; The weighted fusion unit is used to retrieve historical detection results, perform weighted fusion on the multiple enhanced feature data and the historical detection results, and generate real-time anomaly detection results, wherein the real-time anomaly detection results include anomaly type labels and confidence scores.

4. The battery pack thermal runaway risk identification and warning system according to claim 3, characterized in that: Prediction module, including: A selection unit, configured to dynamically select an input feature subset based on the anomaly detection result to construct a multi-dimensional time series input matrix for a target time window; A thermal runaway prediction unit, used to perform thermal runaway prediction on the multi-dimensional time series input matrix to obtain a thermal state prediction data set, wherein the thermal state prediction data set includes a temperature change rate, a gas production rate, and a deformation acceleration prediction value; A first state evolution unit, configured to perform state evolution based on the temperature change rate of the battery pack to obtain spatial temperature distribution data; A second state evolution unit, used to perform state evolution based on the gas production rate of the battery pack to obtain gas diffusion path data; a third state evolution unit, configured to perform state evolution based on the deformation acceleration prediction value of the battery pack to obtain mechanical deformation vector data; The correlation and integration unit is used to correlate and integrate the spatial temperature distribution data, the gas diffusion path data, and the mechanical deformation vector data to draw the thermal state evolution trend diagram.

5. The battery pack thermal runaway risk identification and warning system according to claim 4, characterized in that: Prediction module, including: a data screening unit, configured to screen the operating parameter data set according to the abnormal type label to generate a first feature subset; a weighting unit, configured to weight the first feature subset based on the confidence score to generate a second feature subset; A data extraction unit, used to extract the sliding mean, standard deviation, and kurtosis coefficient of the second feature subset within a target time window, and construct a multidimensional statistical feature matrix; The fusion unit is used to fuse the multidimensional statistical feature matrix with the electrochemical mechanism simulation data at the feature level to generate the multidimensional time series input matrix.

6. The battery pack thermal runaway risk identification and warning system according to claim 1, characterized in that: Assessment modules include: A multidimensional decomposition unit, used for multidimensionally decomposing the thermal state evolution trend diagram to obtain multidimensional data, wherein the multidimensional data includes heat accumulation rate data, heat diffusion gradient data, and thermal runaway trigger probability data; A dynamic adjustment unit, used to dynamically adjust the weight coefficients of the heat accumulation rate data, the heat diffusion gradient data, and the thermal runaway trigger probability data according to the health status parameters of the battery pack to generate a comprehensive risk score; The risk determination unit is used to determine the risk level according to the comprehensive risk score to obtain the thermal runaway risk level.

7. The battery pack thermal runaway risk identification and warning system according to claim 6, characterized in that: Assessment modules include: A traversal unit, used for constructing a risk level quantification matrix based on the comprehensive risk score, and traversing the risk level quantification matrix to extract a thermal diffusion factor, an electrochemical instability index, and a deformation energy accumulation value of a battery pack; A threshold setting unit, used to set a multi-level risk threshold, wherein the multi-level risk threshold includes a first risk threshold, a second risk threshold, and a third risk threshold; A first determination unit, configured to determine that the risk is level I when the thermal diffusion factor exceeds the first risk threshold and the electrochemical instability index is lower than the second risk threshold; A second determination unit, configured to determine that the risk is level II when the thermal diffusion factor exceeds the first risk threshold, the electrochemical instability index exceeds the second risk threshold, and the deformation energy accumulation value is lower than the third risk threshold; The third determination unit is configured to determine that the risk is level III when the thermal diffusion factor exceeds the first risk threshold, the electrochemical instability index exceeds the second risk threshold, and the deformation energy accumulation value exceeds the third risk threshold.

8. The battery pack thermal runaway risk identification and warning system according to claim 7, characterized in that: Assessment modules include: A dimension analysis unit, used to analyze the matrix dimension of the risk level quantification matrix to determine the thermodynamic dimension, the electrochemical dimension and the mechanical dimension; A first calculation unit, configured to calculate, based on the thermodynamic dimension and according to the heat diffusion gradient data, the product of the maximum temperature difference between adjacent monomers in the temperature field and the heat conduction rate as the heat diffusion factor; A second calculation unit is used to obtain the electrochemical instability index based on the electrochemical dimension, based on the thermal runaway trigger probability data and the voltage drop rate calculation of the gas concentration change curve; The third calculation unit is used to analyze the stress concentration area of ​​the battery pack shell based on the mechanical dimension through the deformation vector field combined with the heat accumulation rate data to calculate the deformation energy accumulation value.

9. The battery pack thermal runaway risk identification and warning system according to claim 7, characterized in that: Assessment modules include: A fourth calculation unit is used to calculate the three-dimensional risk diffusion of the battery pack through the heat conduction equation according to the real-time temperature distribution data to determine the thermal runaway propagation path; A first triggering unit is used to trigger a first warning signal when the thermal runaway risk level is level I risk, locate the abnormal position coordinates according to the thermal runaway propagation path, and generate a warning message; A second trigger unit is used to trigger a second warning signal when the thermal runaway risk level is level II risk, dynamically adjust the liquid cooling system according to the thermal runaway propagation path, generate a liquid cooling control instruction, and distribute the flow to the cooling pipeline of the liquid cooling system; The third trigger unit is used to trigger a third warning signal when the thermal runaway risk level is level III risk, activate a directional fire extinguishing device based on the thermal runaway propagation path, and generate an emergency pressure relief channel.

10. A method for identifying and warning the risk of thermal runaway of a battery pack, characterized in that: The method is used to implement the battery pack thermal runaway risk identification and warning system according to any one of claims 1 to 9, and the method comprises: Collecting the operating parameter data set of the battery pack in real time through a multimodal sensor network; Performing hybrid anomaly detection on the battery pack based on the operating parameter data set to obtain a real-time anomaly detection result; Predicting the thermal state of the battery pack according to the real-time abnormality detection result and constructing a thermal state evolution trend diagram; A thermal runaway risk assessment is performed based on the thermal state evolution trend diagram, a thermal runaway risk level is determined, a multi-level warning signal is triggered according to the thermal runaway risk level, and a thermal runaway suppression strategy is generated.

Images