An evaluation method for lithium-ion battery thermal runaway hazard
By combining FDS simulation data and experimental data, and using the analytic hierarchy process (AHP) and entropy weight method to construct a multi-index evaluation system, the problems of single evaluation dimension, strong subjectivity, and insufficient dynamic monitoring in the evaluation of thermal runaway of lithium-ion batteries are solved. This enables a comprehensive, scientific, and quantitative evaluation of thermal runaway of lithium-ion batteries, and provides a quantitative basis for safety design and emergency disaster relief.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- JIANGSU UNIV
- Filing Date
- 2026-03-13
- Publication Date
- 2026-07-10
AI Technical Summary
Existing methods for assessing the hazards of thermal runaway in lithium-ion batteries suffer from limitations such as a single assessment dimension, strong subjectivity, unreasonable weight allocation, insufficient dynamic monitoring, and reliance on physical experiments. These limitations result in poor accuracy and consistency of assessment results, high costs, and potential safety hazards.
By integrating FDS simulation data and experimental data, and combining the analytic hierarchy process (AHP) and entropy weight method, a multi-index evaluation system is constructed. Through the allocation of subjective and objective weights, a dynamic quantitative comprehensive assessment of the risks of thermal runaway, toxic gas emission, and explosion of lithium-ion batteries is achieved. This includes formulating multiple risk indicators, establishing a database, monitoring real-time data, and conducting quantitative assessments.
It enables a comprehensive, scientific, and quantitative assessment of the hazards of thermal runaway in lithium-ion batteries, taking into account both expert experience and objective data, reducing subjective bias, lowering assessment costs, and providing a quantitative basis for precise safety design and emergency response plans.
Smart Images
Figure CN122364693A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of lithium-ion battery energy storage technology, and in particular to an assessment method for the hazards of thermal runaway in lithium-ion batteries. Background Technology
[0002] Currently, lithium-ion batteries are widely used due to their high energy density and long cycle life. However, lithium-ion batteries are prone to thermal runaway under abnormal conditions such as overcharging, over-discharging, short circuits, or high temperatures, which can lead to large amounts of heat release, toxic gas emissions, fires, or even explosions, seriously threatening personal and property safety.
[0003] Existing methods for assessing the hazard of thermal runaway in lithium-ion batteries have the following shortcomings: (1) Single assessment dimension: Most assessments only focus on a single hazard of thermal runaway (such as explosion or flame), lacking a comprehensive assessment of the entire hazard chain of heat release, toxic gas, and explosion; (2) Highly subjective: It relies on expert experience for qualitative judgment, lacks scientific quantitative analysis methods, and has poor accuracy and consistency in evaluation results; (3) Unreasonable weight allocation: In existing assessment methods, the weight allocation of each risk indicator often relies solely on the analytic hierarchy process (AHP) or expert scoring. The weight allocation is entirely based on the subjective experience of the assessors, ignoring the objective information carried by the monitoring data itself (i.e., the dispersion and variability of the data), which makes the assessment results susceptible to human interference and lacks a scientific and rigorous quantitative basis; (4) Insufficient dynamic monitoring: lack of real-time analysis of dynamic data such as flame characteristics during the thermal runaway of lithium-ion batteries; (5) Heavy reliance on physical experiments, high data acquisition costs and unpredictability. Existing evaluation methods largely rely on destructive physical experiments. However, due to the randomness of battery thermal runaway processes and the significant impact of experimental environments (such as the size of enclosed chambers and ventilation conditions) on the results, experimental data has poor repeatability, high acquisition costs, and safety hazards. There is a lack of a digital evaluation method based on numerical simulation (such as FDS software) to replace or supplement high-risk physical testing.
[0004] In summary, the applicant proposes an assessment method for the hazards of thermal runaway in lithium-ion batteries. Summary of the Invention
[0005] To address the technical problems existing in the background art, this invention proposes an assessment method and system for the thermal runaway hazards of lithium-ion batteries that integrates FDS simulation data and experimental data, takes into account both subjective and objective weight allocation, and can cover multiple hazard forms. By constructing a scientific multi-index assessment system and a weight allocation model that combines subjective and objective factors, it achieves a dynamic quantitative comprehensive assessment of the thermal runaway heat release, toxic gas emission, and explosion risks of lithium-ion batteries, accurately predicts the evolution of hazard levels, and provides a quantitative basis for formulating optimal safety designs and emergency disaster relief plans.
[0006] This invention proposes an assessment method for the hazards of thermal runaway in lithium-ion batteries, comprising: S1. Develop multiple risk indicators to assess the thermal runaway process of lithium-ion batteries, and calculate the combined weights of each risk indicator based on the analytic hierarchy process and the entropy weight method. S2. A database is pre-established, which stores basic data corresponding to the risk indicators during the complete evolution of thermal runaway of lithium-ion batteries; S3. Monitor the thermal runaway process of lithium-ion batteries and obtain real-time data of at least some risk indicators; S4. Retrieve basic data corresponding to the current thermal runaway evolution stage from the database, and use it together with the real-time data as the input value of the risk indicator. Based on the combined weight and the input value, quantitatively assess the hazards of thermal runaway of lithium-ion batteries.
[0007] Furthermore, the risk indicators include heat release risk indicators, toxic gas emission risk indicators, and explosion risk indicators.
[0008] Furthermore, the heat release risk index includes at least one of heat release rate, maximum temperature, and total heat release; the toxic gas emission risk index includes at least one of CO concentration and HF concentration; and the explosion risk index includes at least one of mixed gas explosion limit, maximum explosion pressure, and pressure rise rate.
[0009] Furthermore, the basic data includes at least one of the following: simulation data obtained by simulating the thermal runaway process, physical data obtained by actual thermal runaway measurements, and visual data obtained by acquiring flame images of the thermal runaway process.
[0010] Furthermore, the simulation data is acquired through an FDS simulation system, which is used to simulate the thermal runaway process of lithium-ion batteries and outputs simulation data including heat release rate, temperature field distribution, gas concentration and pressure changes.
[0011] Furthermore, the physical data is obtained through thermal runaway tests of lithium-ion batteries. These tests include triggering thermal runaway of lithium-ion batteries by external heating or needle penetration, collecting measured data including temperature, pressure, and gas concentration, and obtaining the maximum explosion pressure and pressure rise rate through closed-cell tests.
[0012] Furthermore, the visual data is acquired through an industrial camera, which captures flame images and identifies flame features and inverts the heat release rate based on a deep learning model.
[0013] Furthermore, the calculation of the combined weights of each risk indicator based on the analytic hierarchy process and the entropy weight method includes: The subjective weights of each risk indicator were determined based on expert experience using the analytic hierarchy process. The objective weights of each risk indicator are determined using the entropy weight method based on the degree of dispersion of the basic or real-time data. The combined weights of each risk indicator are obtained by taking the minimum total deviation between subjective and objective weights as the objective function.
[0014] Furthermore, the quantitative assessment of the thermal runaway hazard of lithium-ion batteries based on the combined weights and input values includes: Determine the optimal combination weights of each risk indicator based on the current stage of thermal runaway evolution; The input values of each risk indicator are normalized. Calculate the comprehensive risk coefficient :
[0015] in For the first The optimal combination weights of each risk indicator For the first The normalized input values corresponding to each risk indicator; The comprehensive risk coefficient is determined based on the preset risk level. Conduct risk assessment.
[0016] This invention achieves a quantitative assessment of the hazards of thermal runaway in lithium-ion batteries by combining the analytic hierarchy process (AHP) and the entropy weighting method to determine the combined weights of risk indicators and integrating pre-established basic data and real-time monitoring data as evaluation inputs. This invention balances expert experience with the objective laws of data, avoiding the subjective bias of single weighting methods, and compensates for the deficiencies of real-time monitoring data through data fusion, providing strong support for the dynamic assessment of the entire thermal runaway process. Attached Figure Description
[0017] Figure 1This is a flowchart of an assessment method for the hazards of thermal runaway in lithium-ion batteries proposed in this invention. Detailed Implementation
[0018] The technical solution of the present invention will now be described in detail with reference to the accompanying drawings and embodiments. The embodiments are implemented based on the technical solution of the present invention, providing detailed implementation methods and specific operating procedures. Obviously, the described embodiments are only a part of the embodiments of the present invention, not all of them, and the scope of protection of the present invention is not limited to the following embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.
[0019] The coupled hazards of "heat release-toxic gas-explosion" caused by thermal runaway of lithium-ion batteries have become a major safety concern in the industry. Current technologies for assessing thermal runaway primarily rely on independent assessments of a single hazard (e.g., assessing only explosion pressure or jet thermal radiation), failing to achieve comprehensive quantification of the entire hazard chain. Furthermore, the weighting allocation is highly subjective and lacks dynamic assessment capabilities. The applicant proposes the following solution to address these existing problems.
[0020] See Figure 1 This invention proposes an assessment method for the hazards of thermal runaway in lithium-ion batteries, comprising: S1. Develop multiple risk indicators to assess the thermal runaway process of lithium-ion batteries, and calculate the combined weights of each risk indicator based on the analytic hierarchy process and the entropy weight method. S2. A database is pre-established, which stores basic data corresponding to the risk indicators during the complete evolution of thermal runaway of lithium-ion batteries; S3. Monitor the thermal runaway process of lithium-ion batteries and obtain real-time data of at least some risk indicators; S4. Retrieve basic data corresponding to the current thermal runaway evolution stage from the database, and use it together with the real-time data as the input value of the risk indicator. Based on the combined weight and the input value, quantitatively assess the hazards of thermal runaway of lithium-ion batteries.
[0021] In a specific embodiment, the Analytic Hierarchy Process (AHP) is used to establish a three-level quantitative assessment system to obtain multi-dimensional risk indicators, thereby overcoming the limitation of existing technologies with a single assessment dimension, as detailed below: (1) Primary indicator (target layer): Total hazards of thermal runaway of lithium-ion batteries.
[0022] (2) Secondary indicators (criteria level): heat release risk indicators, toxic gas emission risk indicators, and explosion risk indicators.
[0023] (3) Third-level indicators (parameter layer): (a) Heat release risk indicators: at least one of the following: heat release rate (HRR), maximum temperature, and total heat release; (b) Risk indicators for toxic gas emissions: at least one of CO (carbon monoxide) concentration and HF (hydrogen fluoride) concentration; (c) Explosion risk indicators: at least one of the following: gas mixture explosion limit (LEL), maximum explosion pressure (MEP), and rate of pressure rise (unit: dP / dt).
[0024] Compared to existing technologies that only target a single type of disaster (such as assessing only explosion pressure or only flame radiation), this invention adopts a multi-dimensional risk assessment system that includes heat release rate, toxic gas concentration (CO / HF), explosion limits, and pressure rise rate. This system not only covers the entire life cycle hazards of thermal runaway but also comprehensively captures the complete evolution process of battery thermal runaway from initial gas production, mid-stage flame propagation to late-stage explosion, avoiding risk omissions due to missing assessment dimensions. Furthermore, it allows for flexible adjustment of indicator weights according to application scenarios, taking into account the differentiated safety needs of different scenarios such as passenger vehicles and energy storage power stations.
[0025] In a specific embodiment, to address the problem of excessive subjectivity or insufficient objectivity in weight allocation in the prior art, step S1 calculates the combined weights of each risk indicator based on the analytic hierarchy process (AHP) and the entropy weight method, including: S11. The subjective weights of each risk indicator are determined based on expert experience using the analytic hierarchy process.
[0026] Specifically, eight risk indicators were selected: lithium-ion battery heat release rate, maximum temperature, total heat release, CO concentration, HF concentration, mixed gas explosion limits, maximum explosion pressure, and pressure rise rate. The data acquisition method for each risk indicator can be flexibly selected for different application scenarios.
[0027] Using Saaty's 1-9 scale, experts in battery safety, fire protection, and energy storage were invited to conduct pairwise comparisons of the indicators to construct a judgment matrix.
[0028] (1) For the criteria layer indicators, the judgment matrix is constructed as follows:
[0029] a. A judgment matrix is constructed for the three risk indicators: heat release rate, maximum temperature, and total heat release, as follows:
[0030] b. For the sub-indicators of toxic gas risk (CO concentration, HF concentration), the judgment matrix is constructed as follows:
[0031] c. For the sub-indicators of explosion risk (explosion limit of mixed gas, maximum explosion pressure, and pressure rise rate), construct the following judgment matrix:
[0032] (2) Calculate the weights of the above indicators and, after consistency testing, obtain the combined weights of the indicator layer relative to the target layer as follows:
[0033] S12. Due to the limitations of the subjective weights in the Analytic Hierarchy Process (AHP), the entropy weight method is used to determine the objective weights of each risk indicator based on the dispersion of the basic or real-time data. .
[0034] Specifically, the multi-source heterogeneous data is transformed to the [0,1] interval using range normalization:
[0035] in, Representing the Data series of risk indicators, For the first The first risk indicator in the data sequence One data point, The first range after standardization transformation The first of the risk indicators Data points.
[0036] Information entropy The overall information content of an indicator can be reflected by calculating its dispersion through information entropy and assigning different weights accordingly.
[0037] Where n represents the amount of data; Indicates the first The data in the first The proportion of each risk indicator k is an adjustment coefficient used to ensure the information entropy value. The calculation result is between 0 and 1. .
[0038] Entropy weight of the j-th risk indicator Calculate using the following formula:
[0039] Where m represents the total number of risk indicators.
[0040] S13. Using the minimum total deviation between subjective weights and objective weights as the objective function, the combined weights of each risk indicator are obtained by solving the problem.
[0041] Specifically, an objective function based on the Lagrange multiplier method that minimizes the total deviation is established. Subjective weights of AHP Objective weights of the entropy weight method Perform optimization and combination to achieve constrained optimization and weighted combination:
[0042] The constraints are:
[0043] in, For the first The weighted indicators obtained by optimizing each risk indicator Indicates the first Subjective weighting of risk indicators using AHP; balance coefficient It can be adjusted according to the application scenario, if If the value is large, the calculation result will emphasize the weights obtained by the entropy weight method; if... If the value is small, the calculation result will emphasize the subjective weights derived from AHP.
[0044] This invention employs a combined weighting structure integrating the Analytic Hierarchy Process (AHP) and the entropy weighting method. On one hand, the AHP leverages the extensive experience of industry experts in lithium battery safety; on the other hand, the entropy weighting method uncovers the objective statistical laws (information entropy) inherent in the thermal runaway process of lithium-ion batteries. Finally, a minimum deviation model dynamically integrates the two methods, overcoming the subjective arbitrariness caused by relying solely on expert scoring in traditional assessments. It also compensates for the shortcomings of relying solely on data statistics, which neglects physical mechanisms, making the allocation of risk weights more scientific and rigorous.
[0045] In a specific embodiment, the basic data in S2 includes at least one of the following: simulation data obtained by simulating the thermal runaway process, physical data obtained by actual thermal runaway measurements, and visual data obtained by acquiring flame images of the thermal runaway process.
[0046] The simulation data is acquired through an FDS simulation system, which is used to simulate the thermal runaway process of lithium-ion batteries. This system constructs a high-precision numerical simulation of the battery thermal runaway process, setting exothermic curves (such as quasi-steady-state thermal runaway exothermic rates based on battery material properties) and gas-generating component ratios (typical components include...) that match the actual thermal runaway process. , , , (etc.) to achieve multiphysics simulation of the entire life cycle of thermal runaway. The simulation process outputs time-series data with a time step of 0.5s, including core parameters such as heat release rate, three-dimensional temperature field distribution, multi-component gas concentration, and battery compartment pressure changes, which can provide data support for risk assessment.
[0047] The physical data was obtained through thermal runaway tests on lithium-ion batteries. These tests included triggering thermal runaway by external heating or needle penetration, real-time monitoring of the maximum temperature of the lithium-ion battery and the environment using a surface thermocouple array, tracking pressure changes within a confined space using a high-frequency pressure sensor, and online measurement of toxic gas concentrations such as CO and HF using a Fourier transform infrared spectroscopy or electrochemical sensors. In the gas safety testing phase, the lower explosive limit (LEL) of the gas mixture was calculated according to Le Chatelier's principle; and the maximum explosion pressure (MEP) and pressure rise rate (dP / dt) were obtained through closed-bulk test.
[0048] The visual data is acquired through an industrial camera, which captures flame images and identifies flame features, such as flame height, area, and combustion morphology, based on a deep learning model (e.g., YOLOv8). This constructs a correlation model between flame characteristics and heat release rate trained on a large amount of experimental data, enabling non-contact inversion estimation of heat release rate through flame visual features.
[0049] This invention introduces machine vision (industrial camera) and an FDS simulation system, solving the problem that traditional contact sensors are easily damaged and lose data under conditions of violent battery thermal runaway or explosion. By utilizing deep learning algorithms to invert the heat release rate from flame images and combining this with FDS simulation to supplement extreme condition data, "non-contact" parameter acquisition is achieved in areas where sensors fail or cannot be installed, ensuring continuous and stable output of risk assessment results even under extremely dangerous conditions.
[0050] Furthermore, by supplementing extreme operating condition data that physical experiments cannot cover through FDS numerical simulation, this invention effectively reduces the number of high-cost, high-risk destructive physical experiments. Simultaneously, the quantified four-level hazard classification (low / medium / high / extremely hazardous) and dynamic judgment mechanism can provide precise quantitative data support for explosion-proof design in the R&D phase and emergency early warning during operation of energy storage power stations or electric vehicles.
[0051] In a specific embodiment, the quantitative assessment of the thermal runaway hazard of lithium-ion batteries based on the combined weights and input values in step S4 includes: Determine the optimal combination weights of each risk indicator based on the current stage of thermal runaway evolution; The input values of each risk indicator are normalized. Calculate the comprehensive risk coefficient :
[0052] in For the first The optimal combination weights of each risk indicator For the first The normalized input values corresponding to each risk indicator; The comprehensive risk coefficient is determined based on the preset risk level. Conduct risk assessment.
[0053] For example, the risk level can be preset as follows: Low risk: ; Medium risk: ; High risk: ; Critically Endangered: .
[0054] The following uses a 280Ah lithium iron phosphate battery module of an energy storage power station as a specific embodiment to illustrate the thermal runaway assessment method of the present invention.
[0055] S1. Develop multiple risk indicators to assess the thermal runaway process of lithium-ion batteries, and calculate the combined weights of each risk indicator based on the analytic hierarchy process and the entropy weight method. S2. Use FDS simulation software to establish a 1:1 numerical model of the battery module, input the thermal property parameters of the battery material, simulate the complete evolution process of thermal runaway under different ventilation conditions (air exchange rate 0-10 times / h), and extract the HRR peak (about 1.2MW), CO concentration distribution and pressure rise curve as the basic data. Obtain physical data of thermal runaway of this type of battery module, including data obtained through thermal runaway tests conducted by the energy storage power station itself, as well as data obtained from existing technical literature or battery manufacturer information; The simulation data and the measured physical data are used to construct a database.
[0056] S3. Monitor the thermal runaway process of lithium-ion batteries by acquiring images at a frequency of 30fps using an industrial camera and calculating the main flame characteristics in real time; at the same time, collect the concentrations of CO and HF in the environment through a gas sensor array.
[0057] S4. Based on the acquired flame images and real-time environmental CO and HF concentrations, combined with database data, determine the current thermal runaway evolution stage. Then, update the objective weights in real-time using the entropy weight method based on the current thermal runaway evolution stage. Set the balance coefficient The optimal combination weights of each risk indicator at the current moment are determined. Combining the basic data stored in the database and the collected real-time data, the input values of each risk indicator are normalized (assuming a time of T=150s), as follows: Machine vision detected the normalized value of the high-temperature region of the flame. The sensor detects the normalized value of CO concentration. The normalized values of the other indicators are approximately 0.5.
[0058] The optimal combination weights calculated based on the current moment. (e.g., CO weight is 0.215, HRR weight is 0.373, etc.); substitute into the comprehensive risk coefficient. The calculation formula yields the comprehensive risk coefficient. It is 0.78.
[0059] Therefore, it is determined that the current state is "extremely dangerous" (0.78>0.75).
[0060] The assessment method of this invention determines the combined weights of risk indicators by combining the analytic hierarchy process (AHP) and the entropy weight method, and integrates pre-established basic data and real-time monitoring data as assessment inputs, thereby achieving a quantitative assessment of the hazards of thermal runaway in lithium-ion batteries. This invention balances expert experience with the objective laws of data, avoiding the subjective bias of single weighting methods, and compensates for the deficiencies of real-time monitoring data through data fusion, providing strong support for the dynamic assessment of the entire thermal runaway process.
[0061] The term "an embodiment" or "embodiment" as used in this invention refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the invention. In the description of this invention, it should be understood that the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the steps or units listed, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or apparatuses.
[0062] This specification provides method operation steps as shown in the embodiments or flowcharts, but based on conventional or non-inventive labor, more or fewer operation steps may be included. The order of steps listed in the embodiments is merely one of many possible execution orders and does not represent the only possible execution order. In actual system or server product execution, the method can be executed in the order shown in the embodiments or drawings, or in parallel (e.g., in a parallel processor or multi-threaded processing environment), or the execution order of steps without timing constraints can be adjusted.
[0063] The preferred embodiments of the present invention have been described in detail above. It should be understood that those skilled in the art can make numerous modifications and variations based on the concept of the present invention without creative effort. Therefore, all technical solutions that can be obtained by those skilled in the art based on the concept of the present invention through logical analysis, reasoning, or limited experimentation on the basis of existing technology should be within the scope of protection defined by the claims.
Claims
1. A method for assessing the hazards of thermal runaway in lithium-ion batteries, characterized in that, include: S1. Develop multiple risk indicators to assess the thermal runaway process of lithium-ion batteries, and calculate the combined weights of each risk indicator based on the analytic hierarchy process and the entropy weight method. S2. A database is pre-established, which stores basic data corresponding to the risk indicators during the complete evolution of thermal runaway of lithium-ion batteries; S3. Monitor the thermal runaway process of lithium-ion batteries and obtain real-time data of at least some risk indicators; S4. Retrieve basic data corresponding to the current thermal runaway evolution stage from the database, and use it together with the real-time data as the input value of the risk indicator. Based on the combined weight and the input value, quantitatively assess the hazards of thermal runaway of lithium-ion batteries.
2. The evaluation method according to claim 1, characterized in that, The risk indicators include heat release risk indicators, toxic gas emission risk indicators, and explosion risk indicators.
3. The evaluation method according to claim 2, characterized in that, The heat release risk indicators include at least one of heat release rate, maximum temperature, and total heat release; the toxic gas emission risk indicators include at least one of CO concentration and HF concentration; and the explosion risk indicators include at least one of mixed gas explosion limit, maximum explosion pressure, and pressurization rate.
4. The evaluation method according to claim 1, characterized in that, The basic data includes at least one of the following: simulation data obtained by simulating the thermal runaway process, physical data obtained by actual thermal runaway measurements, and visual data obtained by acquiring flame images of the thermal runaway process.
5. The evaluation method according to claim 4, characterized in that, The simulation data is acquired through the FDS simulation system, which is used to simulate the thermal runaway process of lithium-ion batteries and outputs simulation data including heat release rate, temperature field distribution, gas concentration and pressure changes.
6. The evaluation method according to claim 4, characterized in that, The physical data was obtained through thermal runaway tests of lithium-ion batteries. The thermal runaway tests included triggering thermal runaway of lithium-ion batteries by external heating or needle penetration, collecting measured data including temperature, pressure, and gas concentration, and obtaining the maximum explosion pressure and pressure rise rate through closed-capacity tests.
7. The evaluation method according to claim 4, characterized in that, The visual data is acquired through an industrial camera, which captures flame images and identifies flame features and inverts the heat release rate based on a deep learning model.
8. The evaluation method according to claim 1, characterized in that, The calculation of the combined weights of each risk indicator based on the analytic hierarchy process (AHP) and entropy weight method includes: The subjective weights of each risk indicator were determined based on expert experience using the analytic hierarchy process (AHP). The objective weights of each risk indicator are determined using the entropy weight method based on the degree of dispersion of the basic or real-time data. The combined weights of each risk indicator are obtained by taking the minimum total deviation between subjective and objective weights as the objective function.
9. The evaluation method according to claim 1, characterized in that, The quantitative assessment of the thermal runaway hazard of lithium-ion batteries based on the combined weights and input values includes: Determine the optimal combination weights of each risk indicator based on the current stage of thermal runaway evolution; The input values of each risk indicator are normalized. Calculate the comprehensive risk coefficient : in For the first The optimal combination weights of each risk indicator For the first The normalized input values corresponding to each risk indicator; The comprehensive risk coefficient is determined based on the preset risk level. Conduct risk assessment.