Lithium ion battery thermal runaway multi-cascade emergency response prevention system

Abstract

The invention relates to the technical field of lithium ion battery safety, and discloses a lithium ion battery thermal runaway multi-cascade emergency response prevention system. The system comprises a thermal signal monitoring module, a thermal diffusion prediction module, an emergency suppression module, a pressure balance module and a multi-mode linkage module. The thermal signal monitoring module obtains multi-dimensional temperature data, identifies abnormity and generates a thermal runaway early warning signal; the thermal diffusion prediction module combines the early warning signal and the battery module structure parameters to calculate the thermal propagation path and rate and generate a prediction result; the emergency suppression module matches a suppression medium according to a prediction result, and calculates a dose putting strategy to form a suppression scheme; the pressure balance module obtains pressure data, analyzes the change rate and distribution, adjusts a pressure release valve strategy in combination with a suppression scheme, and generates a pressure regulation and control instruction; and the multi-mode linkage module integrates data and a scheme, and generates a system-level emergency response instruction based on a pressure regulation and control instruction to realize thermal runaway multi-level joint prevention.

Description

Technical Field

[0001] This invention relates to the field of lithium-ion battery safety technology, specifically to a multi-level joint prevention and emergency response system for thermal runaway in lithium-ion batteries. Background Technology

[0002] With the rapid development of the new energy industry, lithium-ion batteries, with their advantages of high energy density, long cycle life, and no memory effect, are widely used in electric vehicles, energy storage power stations, and portable electronic devices. However, lithium-ion batteries are susceptible to thermal runaway during use due to factors such as internal material properties, external operating environment, and operating conditions. When a battery is subjected to overcharging, over-discharging, short circuits, compression, puncture, or high-temperature environments, violent chemical reactions can easily occur internally, releasing a large amount of heat and causing a rapid increase in temperature. If the heat cannot be controlled in time, it will quickly spread to adjacent batteries, forming a chain reaction and leading to thermal runaway of the entire battery pack.

[0003] Battery thermal runaway not only causes permanent damage to battery performance but can also trigger safety accidents such as smoke, fire, and even explosion, posing a serious threat to personal safety and property. Currently, protective measures against lithium-ion battery thermal runaway mostly focus on monitoring or suppressing single aspects, such as using temperature sensors to monitor battery surface temperature or installing fire extinguishing devices within the battery pack. However, these single-faceted protective methods have significant limitations: relying solely on temperature monitoring is insufficient to comprehensively capture abnormal signals in the early stages of thermal runaway within the battery; often, by the time an anomaly is detected, the thermal runaway has already entered its middle or late stages, missing the optimal intervention window. Traditional fire extinguishing devices lack precise judgment of heat diffusion paths, and their suppression medium deployment strategies are relatively crude, making it difficult to effectively suppress different heat diffusion rates and ranges. Furthermore, they ignore the additional safety hazards caused by the sudden increase in internal pressure during thermal runaway—excessive internal pressure can not only cause the battery pack casing to rupture but may also accelerate the spread of flames and toxic gases, further expanding the scope of the accident.

[0004] In existing protection systems, there is a lack of effective coordination and linkage between functional modules. Temperature monitoring, thermal runaway analysis, suppression execution, and pressure control are all independent processes, failing to form a unified emergency response system. For example, when the temperature monitoring module issues an early warning signal, the suppression module fails to adjust its actions in a timely manner based on the thermal runaway prediction results, and the pressure control module does not dynamically adjust its pressure relief strategy according to pressure changes during the suppression process. This results in low efficiency throughout the emergency response process, making it difficult to effectively curb the spread of thermal runaway.

[0005] In the electric vehicle sector, battery packs, as the core power source, are particularly vulnerable to thermal runaway. During electric vehicle operation, the battery pack operates in a complex and variable environment. Acceleration, deceleration, and hill climbing can cause fluctuations in battery load, further increasing the risk of thermal runaway. In energy storage power stations, the centralized deployment of large-scale battery packs exacerbates the chain reaction of thermal runaway, leading to more profound economic losses and social impacts in the event of an accident. Therefore, current single, decentralized methods of lithium-ion battery thermal runaway protection are insufficient to meet the high safety requirements of practical applications. A multi-level, interconnected emergency response system is needed, capable of multi-stage collaboration, multi-dimensional monitoring, accurate prediction, and dynamic control. This system will enhance the comprehensiveness, timeliness, and effectiveness of lithium-ion battery thermal runaway protection, ensuring the safe and stable operation of lithium-ion batteries in various applications. Summary of the Invention

[0006] The purpose of this invention is to provide a multi-level joint prevention and emergency response system for thermal runaway of lithium-ion batteries, so as to solve the problems mentioned in the background art.

[0007] To achieve the above objectives, the present invention provides a multi-level joint prevention and emergency response system for thermal runaway of lithium-ion batteries, the system comprising:

[0008] The thermal signal monitoring module acquires multi-dimensional temperature sensor data inside the battery pack, identifies abnormal temperature signals and temperature change trends, and generates a battery thermal runaway early warning signal.

[0009] The thermal runaway prediction module calculates the heat propagation path and diffusion rate based on the battery thermal runaway early warning signal and the battery module structural parameters, and generates thermal runaway prediction results.

[0010] The emergency suppression module matches the suppression medium and calculates the inhibitor dosage strategy based on the thermal diffusion prediction results, and generates a thermal runaway suppression scheme.

[0011] The pressure balance module acquires pressure sensor data inside the battery pack, analyzes the rate of pressure change and pressure distribution, adjusts the pressure relief valve action strategy according to the thermal runaway suppression scheme, and generates pressure control commands.

[0012] Based on the pressure control command, the multi-mode linkage module integrates thermal signal data, diffusion prediction results, and suppression schemes to generate system-level emergency response commands.

[0013] Preferably, the battery thermal runaway early warning signal includes an abnormal temperature area identifier, a temperature change acceleration value, and a thermal runaway risk level; the thermal diffusion prediction result includes a predicted heat propagation direction, a heat transfer rate between modules, and an estimated critical diffusion time; the thermal runaway suppression scheme includes the type of suppression medium, the dosage of the medium, and the distribution of the medium's placement location; the pressure control command includes the timing of the pressure relief valve opening, the duration of pressure relief, and the pressure release intensity classification; and the system-level emergency response command includes multi-level response triggering conditions, system action priority ranking, and linkage control timing.

[0014] Preferably, the thermal signal monitoring module includes:

[0015] The temperature signal acquisition submodule acquires data from multiple temperature sensors arranged inside the battery pack, aggregates temperature readings by region, and records the temperature value for each acquisition cycle.

[0016] The temperature change analysis submodule calculates the temperature change difference between adjacent time points based on the temperature value, and determines whether the temperature change exceeds the safety threshold by combining the normal operating temperature range of the battery.

[0017] The thermal risk assessment submodule compares the temperature change difference with the temperature characteristics of historical thermal runaway cases and outputs a battery thermal runaway warning signal.

[0018] Preferably, the thermal diffusion prediction module includes:

[0019] Based on the battery thermal runaway early warning signal, the module thermal conduction analysis submodule calls the thermal conduction coefficient between battery modules and the structural gap parameters to calculate the heat transfer rate between modules.

[0020] The diffusion path simulation submodule simulates possible heat propagation paths based on the heat transfer rate between modules and the battery pack spatial layout data.

[0021] The thermal impact prediction submodule estimates the time it takes for heat to reach critical components based on the possible propagation paths and generates thermal diffusion prediction results.

[0022] Preferably, the emergency suppression module includes:

[0023] Based on the thermal diffusion prediction results, the suppression medium matching submodule queries the medium property database and selects a suppression medium that matches the thermal runaway type.

[0024] The dosage calculation submodule calculates the required amount of suppression medium to be administered based on the heat transfer rate and the range of influence in the heat diffusion prediction results.

[0025] The deployment strategy generation submodule, in conjunction with the internal structure of the battery pack, determines the release points and sequence of the suppression medium, and generates a thermal runaway suppression scheme.

[0026] Preferably, the pressure balancing module includes:

[0027] The pressure data acquisition submodule acquires real-time data from multiple pressure sensors inside the battery pack and records the pressure values ​​and change times by region.

[0028] The pressure trend analysis submodule calculates the rate of pressure rise per unit time based on the real-time data and compares it with the safe pressure range.

[0029] The pressure relief strategy adjustment submodule determines the triggering timing and opening range of the pressure relief valve based on the pressure rise rate and thermal runaway suppression scheme, and generates pressure control commands.

[0030] Preferably, the multi-mode linkage module includes:

[0031] The data integration submodule receives the battery thermal runaway early warning signal, thermal diffusion prediction results, thermal runaway suppression scheme and pressure control command, and aligns the multi-source data according to the time series.

[0032] The response priority calculation submodule assesses the thermal risk level and pressure hazard level of different regions based on the multi-source data, and determines the priority order of response actions;

[0033] The linkage instruction generation submodule generates coordinated operation instructions, including suppression, pressure relief, and alarm, according to the aforementioned priority order, and outputs system-level emergency response instructions.

[0034] Preferably, the thermal signal monitoring module further includes a signal verification submodule, which acquires temperature data from multiple sensors in the same area, compares the consistency of the readings of each sensor, eliminates abnormal readings, and recalculates the temperature change trend.

[0035] Preferably, the heat diffusion prediction module further includes a real-time correction submodule, which receives the latest temperature data, compares it with the predicted diffusion path, and dynamically adjusts the predicted heat transfer rate and diffusion direction.

[0036] Preferably, the multi-mode linkage module further includes a feedback learning sub-module, which records the execution effect data of each emergency response, updates the historical case feature library, and optimizes the thermal risk judgment rules and response strategies.

[0037] Compared with the prior art, the beneficial effects of the present invention are:

[0038] From the perspective of thermal runaway early warning, the thermal signal monitoring module does not rely on a single temperature monitoring point, but rather acquires multi-dimensional temperature sensor data within the battery pack. This multi-dimensional monitoring method can cover temperature changes in different areas and locations within the battery pack. It can not only capture temperature anomalies on the battery surface, but also more sensitively identify temperature change trends in the early stages of thermal runaway within the battery. This allows for earlier detection of thermal runaway risks, resulting in more timely and accurate battery thermal runaway early warning signals. This avoids the problem of traditional single-temperature monitoring missing the optimal intervention opportunity due to its limited monitoring range, thus buying valuable time for the activation of subsequent emergency response measures.

[0039] In terms of thermal runaway control, the thermal runaway prediction module calculates the heat propagation path and diffusion rate based on thermal runaway early warning signals and battery module structural parameters. The generated thermal runaway prediction results clearly show the direction and speed of potential thermal runaway propagation. This design allows the emergency suppression module to move beyond a crude suppression approach and instead target and match suppression media based on accurate thermal runaway prediction results. For example, suitable suppression media can be selected for different types of thermal runaway reactions and battery material characteristics. Simultaneously, the suppression dosage strategy is calculated based on the thermal runaway range and rate, resulting in a more targeted and effective thermal runaway suppression scheme. This scheme can precisely target critical thermal runaway paths, effectively curbing further heat spread and avoiding the poor suppression effect of traditional suppression methods due to a lack of understanding of thermal runaway patterns.

[0040] For internal pressure control of the battery pack, the pressure balancing module fills the gap in traditional protection systems that neglect pressure risks. This module acquires data from internal pressure sensors, analyzes the rate of pressure change and pressure distribution in real time, and dynamically adjusts the pressure relief valve's action strategy based on the thermal runaway suppression scheme. During thermal runaway, the chemical reaction inside the battery generates a large amount of gas, causing a sudden increase in internal pressure. The pressure control commands generated by the pressure balancing module ensure that the pressure relief valve releases pressure at the appropriate time and rate. This avoids problems such as battery pack casing rupture and accelerated diffusion of hazardous substances due to excessive pressure, while also preventing external impurities from entering the battery pack or affecting the effectiveness of the suppression medium due to excessive pressure relief. This achieves precise control of the internal pressure of the battery pack, further enhancing the safety of emergency response.

[0041] The multi-mode linkage module, as the core coordination unit of the system, integrates thermal signal data, diffusion prediction results, and suppression plans, generating system-level emergency response commands based on pressure control instructions. This design breaks down the barriers of independent operation of functional modules in traditional protection systems, achieving deep collaboration among all aspects. For example, when the thermal signal monitoring module detects an abnormal temperature and issues an early warning, the thermal diffusion prediction module quickly calculates the heat propagation path, the emergency suppression module prepares a suppression plan based on the prediction results, and the pressure balance module simultaneously monitors pressure changes. The multi-mode linkage module integrates this information to ensure that the release of the suppression medium and the action of the pressure relief valve are coordinated. While releasing the suppression medium, it adjusts the pressure relief strategy in a timely manner according to pressure changes to avoid sudden pressure increases affecting the suppression effect and to prevent premature pressure relief leading to the loss of the suppression medium. This collaborative linkage mechanism makes the entire emergency response process more orderly and efficient, enabling real-time adjustment of response strategies to the dynamic development of thermal runaway, effectively improving the overall ability to contain thermal runaway.

[0042] In the electric vehicle sector, the system can adapt to complex operating conditions during vehicle operation, monitor battery pack temperature and pressure changes in real time, accurately predict thermal runaway based on battery module structural characteristics, and promptly initiate suppression and pressure relief measures, providing reliable protection for vehicle operation safety. In energy storage power stations, considering the centralized deployment of large-scale battery packs, the system, through multi-dimensional monitoring and accurate prediction, can effectively prevent the chain reaction of thermal runaway and avoid the expansion of the accident scope. Compared with traditional single protection methods, this multi-level joint prevention and emergency response system forms a complete closed-loop protection system from early warning, prediction, suppression to pressure control, comprehensively covering key risk points in the thermal runaway process, significantly improving the safety protection level of lithium-ion batteries during use, and reducing the probability of thermal runaway accidents and the damage caused by such accidents. Attached Figure Description

[0043] Figure 1 This is a timing diagram of the multi-level joint prevention and emergency response system for thermal runaway of lithium-ion batteries described in this invention;

[0044] Figure 2 A flowchart illustrating the operation of the thermal diffusion prediction module;

[0045] Figure 3 A flowchart illustrating the operation of the emergency suppression module;

[0046] Figure 4 A flowchart illustrating the operation of the multi-mode linkage module. Detailed Implementation

[0047] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0048] See also Figure 1 This invention provides a multi-level joint prevention and emergency response system for thermal runaway in lithium-ion batteries. The system includes: a thermal signal monitoring module, a thermal diffusion prediction module, an emergency suppression module, a pressure balancing module, and a multi-mode linkage module. Specific implementation details are as follows:

[0049] The thermal signal monitoring module acquires multi-dimensional temperature sensor data within the battery pack, identifies abnormal temperature signals and temperature change trends, and generates a battery thermal runaway early warning signal. The thermal propagation prediction module, based on the battery thermal runaway early warning signal and combined with battery module structural parameters, calculates the heat propagation path and diffusion rate, generating a thermal runaway prediction result. The emergency suppression module, based on the thermal runaway prediction result, matches the suppression medium and calculates the suppression dosage strategy, generating a thermal runaway suppression plan. The pressure balance module acquires pressure sensor data inside the battery pack, analyzes the pressure change rate and pressure distribution, adjusts the pressure relief valve action strategy according to the thermal runaway suppression plan, and generates a pressure control command. The multi-mode linkage module, based on the pressure control command, integrates thermal signal data, diffusion prediction results, and suppression plan to generate a system-level emergency response command. All modules in the system communicate via a data bus, achieving real-time data exchange and command coordination, ensuring rapid response in the early stages of thermal runaway, effectively suppressing thermal propagation and maintaining system safety.

[0050] Example 1: The generation of battery thermal runaway early warning signals begins with the continuous acquisition and analysis of multi-dimensional temperature sensor data within the battery pack. The system deploys multiple high-precision temperature sensor arrays at key locations inside the battery pack, such as cell gaps, module connections, and the inner wall of the outer casing. These sensors acquire temperature data at millisecond-level frequencies, and the data is transmitted to the central processing unit for real-time analysis. Abnormal temperature zone identification is achieved by spatially gridding the acquired temperature data. The system divides the interior of the battery pack into several virtual regions, each containing a set of sensor nodes. When the temperature reading in a certain region continuously deviates from its historical normal operating temperature range and forms a significant temperature gradient with adjacent regions, the system marks that region as an abnormal temperature zone and assigns it a unique zone identifier. This identifier includes location coordinates, zone type, and preliminary risk assessment information. The calculation of the temperature change acceleration value is accomplished by differentiating the continuous time-series temperature data. The system not only calculates the instantaneous rate of temperature change but, more importantly, the rate of change of the rate of temperature change, i.e., acceleration. This value reflects the rapidity of thermal runaway. A slowly rising temperature curve may originate from environmental changes or normal load fluctuations, while a temperature rise with a high acceleration value strongly suggests the onset of an internal short circuit or exothermic reaction. The system compares the calculated acceleration values ​​with preset multi-level thresholds, which are set based on the battery chemistry, structure, and historical operating data. The thermal runaway risk level is a comprehensive assessment based on the identification of abnormal temperature regions and the acceleration value of temperature changes. The system establishes a risk assessment matrix, combining the degree, range, and drasticness of temperature anomalies to output three risk levels: low, medium, and high. Each level corresponds to different response plans and activation conditions.

[0051] The core of thermal runaway prediction lies in the quantitative analysis and prediction of the heat propagation process. The prediction of heat propagation direction not only considers simple heat conduction principles but also integrates precise modeling of the actual structural layout between battery modules. The system's built-in digital twin model of the battery pack includes the precise geometry, material properties, relative position, and connection method of each module. When a region is identified as the origin of thermal runaway, the system simulates the most probable heat propagation path based on the point's location, current temperature, and the thermal conduction characteristics between modules. These paths typically follow the direction of least thermal resistance, such as through metal connectors, cooling pipes, or directly through the gaps between adjacent cells. The calculation of the heat transfer rate between modules is a dynamic process. Based on real-time temperature data and parameters such as material thermal conductivity, specific heat capacity, and contact thermal resistance, the system calculates the rate at which heat is transferred from one module to an adjacent module. This rate is not a fixed value but is dynamically updated as temperature increases and material properties change. The system continuously monitors the trend of the transfer rate, as a sudden increase in the rate may indicate that thermal runaway has entered a more severe stage. The critical diffusion time estimation is calculated based on the current heat transfer rate and the distance to each critical component. Critical components include adjacent modules, the battery management system unit, high-voltage connectors, and the package structure. The system calculates the time required for heat to reach each critical component and takes the shortest time as the critical diffusion time. This time parameter provides a clear time window for emergency response, directly affecting the timing and intensity of containment measures.

[0052] Developing a thermal runaway suppression scheme is a highly targeted and precisely calculated process. The selection of the suppression medium relies on a pre-established database of medium properties. This database details the physicochemical properties of various suppression media, such as perfluorohexanone, aerosols, and fine water mist, including boiling point, endothermic capacity, electrical insulation, and compatibility with lithium-ion battery materials. Based on the thermal runaway type determined by the thermal runaway warning signal (e.g., primarily caused by an internal short circuit within the cell, or accompanied by electrolyte combustion) and the temperature level and diffusion range indicated in the thermal diffusion prediction results, the system selects the most suitable medium type from the database. The calculation of the medium dosage is a precise estimation process based on energy balance. The system estimates the total heat to be dissipated based on the impact range (number or volume of cells involved), current temperature, and heating rate in the thermal diffusion prediction results. Then, based on the endothermic capacity per unit mass or per unit volume of the selected medium, the total amount of medium to be added is calculated. The distribution of the addition locations is designed to maximize suppression efficiency. The system combines a three-dimensional model of the battery pack's internal structure with predicted heat propagation paths to determine a set of optimal release nozzle or release point locations. The strategy aims to rapidly create an effective suppression concentration around the heat source and block the predicted heat propagation paths. The release sequence typically follows a principle of from the inside out and from the core to the periphery, prioritizing the protection of critical components and preventing large-scale diffusion.

[0053] The generation of pressure control commands is based on real-time monitoring and forward-looking analysis of the internal pressure status of the battery pack. Determining the timing of the pressure relief valve opening requires balancing internal pressure release with external environmental safety. The system continuously monitors the pressure rise rate and absolute pressure value. When the pressure rise rate exceeds a safety threshold and the absolute pressure approaches the pack's design limit, the system calculates the optimal opening time for the pressure relief valve. This timing needs to consider the effect of inhibitor administration, as the administration of certain inhibitors may temporarily alter the internal pressure status. The pressure relief duration is not a fixed value but a dynamically adjusted process. After the pressure relief valve opens, the system continuously monitors the internal pressure decline curve. The command controls the pressure relief valve to remain open until the pressure stabilizes within a preset safe range and the pressure rise rate returns to zero. Pressure release intensity grading is achieved by controlling the opening area or opening time of the pressure relief valve. The system categorizes the release intensity into multiple levels based on the real-time pressure hazard level. Minor overpressure may only require a small amount of pressure relief, while a rapid, high-risk pressure rise will trigger the maximum pressure relief command to reduce internal pressure as quickly as possible and prevent pack rupture.

[0054] System-level emergency response commands are the final output of multi-module information integration and decision-making collaboration. Multi-level response trigger conditions are a series of predefined logical judgment rules based on multi-parameter combinations. These conditions integrate information from multiple dimensions, such as thermal risk level, pressure hazard level, and critical diffusion time. For example, when the thermal risk level is "high" and the pressure rise rate simultaneously exceeds a threshold, the highest level of emergency response will be triggered. System action priority ranking is accomplished through a dynamic evaluation algorithm. This algorithm performs real-time risk scanning on all areas and components within the battery pack that require protection, calculates a "risk score" for each object, and determines the priority order of actions based on the score. Typically, directly suppressing out-of-control points, protecting adjacent high-risk modules, and ensuring pressure balance receive the highest priority. The linkage control sequence is the final generated sequence of operation commands that can be directly executed by each actuator. This sequence, presented in a timeline format, precisely specifies at what time, which pressure relief valve opens at what magnitude, what location, what dosage of which inhibitor is administered, and how it coordinates with the system's alarm and status reporting functions. This sequence ensures that all emergency actions are executed in an orderly, coordinated and efficient manner, avoiding reduced response effectiveness due to conflicting actions or disordered order.

[0055] Example 2: See Figure 2 The operation of the thermal signal monitoring module begins with the temperature signal acquisition submodule. This submodule is responsible for managing and analyzing a multi-point temperature sensor network deployed within the battery pack. These sensors are embedded in a grid pattern between battery modules, on the surface of the cells, and on the inner wall of the battery pack casing, forming a three-dimensional sensing capability of the internal temperature field of the battery. Each sensor continuously measures the temperature value at its location at a fixed acquisition cycle, for example, 10 times per second. The raw temperature readings acquired are transmitted to the processing unit via a data bus. The temperature signal acquisition submodule does not simply record the data; instead, it first preprocesses the data, including filtering to eliminate transient interference signals and grouping and aggregating readings from multiple sensors in the same physical area. Grouping is based on the physical location encoding of the sensors, with each area group containing several adjacent sensors. This submodule records the temperature value sequence of each area group within each acquisition cycle, attaching precise timestamps and location tags, providing a structured data foundation for subsequent analysis.

[0056] The temperature change analysis submodule receives aggregated temperature data from the acquisition submodule. Its core function is to calculate the difference in temperature values ​​between adjacent time points, i.e., the rate of temperature change. This calculation is performed independently for each region group, comparing the average temperature at the current moment with the average temperature at the previous moment to obtain a difference reflecting the rate of temperature change in that region. However, the simple difference in temperature change is insufficient to accurately identify anomalies. This submodule stores temperature range curves for this type of battery under various normal operating conditions, including the allowable temperature fluctuation range under different states such as charging, discharging, and resting. The calculated temperature change difference is compared in real time with the safety threshold under the current operating conditions. The safety threshold is not a fixed value but a dynamic range that adaptively adjusts based on the battery's current operating state (such as charging / discharging current and ambient temperature). When the temperature change difference of a certain region group continuously and significantly exceeds its corresponding dynamic safety threshold, this submodule marks it as a potential abnormal event and marks the data stream of that region as high priority, sending it to the next submodule for in-depth analysis.

[0057] The thermal risk assessment submodule is the decision-making stage of the thermal signal monitoring module. It receives labeled data from the temperature change analysis submodule, focusing on temperature change differences exceeding safety thresholds. This submodule accesses a historical thermal runaway case feature database, which stores numerous temperature evolution patterns recorded in past experiments or actual operations that ultimately led to thermal runaway. These patterns include not only absolute temperature values ​​but, more importantly, the curve shape of temperature changes over time, acceleration characteristics, and correlations between different regions. The thermal risk assessment submodule performs pattern matching and similarity calculations between the currently calculated real-time temperature change trend, including its rate of change, acceleration, and spatial distribution pattern, and the historical case features in the database. Through this comparative analysis, the submodule assesses the likelihood of the current abnormal temperature signal developing into thermal runaway and ultimately outputs a structured battery thermal runaway warning signal. This signal not only includes a binary judgment of whether a risk exists but also the specific identification of the abnormal temperature region, the calculated temperature change acceleration value, and a quantified thermal runaway risk level.

[0058] The thermal runaway prediction module is activated upon receiving a battery thermal runaway warning signal. Its first submodule, the module thermal conduction analysis submodule, immediately begins operation. This submodule calls upon pre-stored detailed parameters of the battery modules, including the thermal conductivity of the materials used between modules, the physical gap dimensions between modules, the contact area, and the presence of thermal interface materials. Combining the abnormal heat source location and temperature data provided in the warning signal, this submodule applies thermal conduction theory to calculate the rate at which heat is transferred from the abnormal area to adjacent modules. This calculation considers multi-dimensional heat transfer paths, including conduction through connecting copper busbars, convection through air gaps, and radiative heat transfer, ultimately synthesizing a key indicator reflecting the speed of heat propagation: the inter-module thermal transfer rate.

[0059] The diffusion path simulation submodule goes a step further, utilizing the heat transfer rate calculated by the module thermal conduction analysis submodule and combining it with high-precision three-dimensional spatial layout data of the battery pack. This spatial layout data defines the precise coordinates and geometry of each battery module, structural component, and cooling pipe within the pack. This submodule uses simulation algorithms to predict the most likely heat propagation paths. The simulation process calculates multiple possible paths and, based on the principle of minimum thermal resistance and highest energy transfer efficiency, ranks these paths by probability, identifying the primary and secondary heat propagation paths, thus outlining a macroscopic picture of heat diffusion.

[0060] The thermal impact prediction submodule is the final output of the prediction module. Based on the possible propagation paths provided by the diffusion path simulation submodule, it performs time-dimensional calculations. This submodule identifies key components within the battery pack, such as the battery management unit, high-voltage terminals, and other undamaged modules, and calculates the time required for heat to reach these key components from the current abnormal heat source point along the predicted heat propagation path. This estimation process generates a critical diffusion time estimate, providing crucial time window information for subsequent emergency response. Finally, the thermal diffusion prediction module integrates all these analytical results—heat propagation direction prediction, inter-module heat transfer rate, and critical diffusion time estimate—to generate a comprehensive thermal diffusion prediction result, providing a scientific basis for the emergency suppression module's decision-making.

[0061] The entire process embodies a progressive analysis approach, from data acquisition to local anomaly detection and then to global situation prediction, enabling the system to make relatively accurate judgments about the possible direction and speed of thermal runaway in its early stages.

[0062] Example 3: See Figure 3The activation of the emergency suppression module relies on the thermal runaway prediction results generated by the thermal runaway prediction module. The primary task of this module is to select a suitable suppression medium, a task performed by the suppression medium matching submodule. This submodule has a built-in database containing the properties of various suppression media, detailing the physicochemical parameters of media such as perfluorohexanone, heptafluoropropane, fine water mist, inert gases, and novel composite inhibitors. These parameters include, but are not limited to, the boiling point, latent heat of vaporization, specific heat capacity, electrical insulation strength, chemical stability, and reaction characteristics with lithium-ion battery components. The matching process is not a simple query but a multi-objective decision-making process. The thermal runaway prediction results received by the submodule include the heat source temperature, the predicted maximum temperature, the heat transfer rate, and potentially involved substances. The submodule calculates the matching degree between these input parameters and the medium properties in the database, prioritizing media types with high heat absorption efficiency, good insulation performance, and the best chemical compatibility with the current thermal runaway scenario. For example, for thermal runaway inside the battery cell at high temperatures and potentially accompanied by electrical short circuits, perfluorohexanone, which has strong vaporization heat absorption capacity and good insulation, may be the preferred choice; while for cases where there is open flame and the area is large, fine water mist or other composite strategies that have both suppression and cooling capabilities may need to be considered.

[0063] Once the type of suppressant is determined, the dosage calculation submodule begins operation. The core function of this submodule is to accurately calculate the required dosage of suppressant. The calculation is based on a thermal balance model designed to counteract excess heat generated by thermal runaway and cool relevant components. The calculation considers multiple variables: the thermal influence range provided in the thermal diffusion prediction results, the currently monitored temperature, the rate of temperature rise, and the characteristic parameters of the selected medium. The calculation takes into account all endothermic processes as the medium progresses from its storage state to its effective function, including the medium's temperature rise, phase change, and further temperature rise of gaseous products. The dosage calculation needs to be sufficiently accurate to avoid underdosing leading to suppression failure, or overdosing resulting in unnecessary waste or potential side effects. This submodule outputs a calculated, exact amount of medium required, typically expressed in units of mass or volume.

[0064] The deployment strategy generation submodule receives media demand information from the dosage calculation submodule and diffusion path information from the heat diffusion prediction module. Its task is to transform the abstract dosage into a concrete, actionable deployment plan. This submodule accesses a detailed 3D structural model of the battery pack, which annotates the location, type, and coverage of all potential inhibitor release points (nozzles). Strategy generation first requires determining the selection of release points. Based on the heat source location and the predicted main heat propagation path, the submodule calculates an optimal combination of release points designed to quickly surround the heat source and block the heat diffusion path. For example, it might select several points closest to the heat source for direct, concentrated suppression, while simultaneously deploying points along the critical downstream path of the heat flow for interception. Secondly, the submodule needs to determine the deployment order and timing. Deployment does not occur simultaneously at all selected points. The strategy may require the points closest to the heat source to activate first to quickly suppress the core reaction, followed by the sequential activation of peripheral points to build a complete suppression barrier, with the time intervals set according to the heat transfer rate. Finally, this submodule generates a detailed thermal runaway suppression plan, which clearly includes the selected suppression medium type, the calculated precise dosage, the selected specific release point distribution, and the release sequence and timing instructions for each point.

[0065] The pressure balancing module and the emergency suppression module operate in parallel, focusing on managing pressure changes within the battery pack during thermal runaway and suppression. Their operation is based on an independent network of pressure sensors. The pressure data acquisition submodule is responsible for acquiring real-time data from pressure sensors distributed throughout the battery pack. These sensors collect pressure readings at a high frequency. The submodule performs preliminary data processing, including filtering to remove noise, grouping the data by sensor region, and recording the pressure value and its changes for each region at each time point.

[0066] The pressure trend analysis submodule performs in-depth analysis of the collected pressure data. It calculates the pressure change per unit time, i.e., the rate of pressure increase or decrease. This rate of pressure change is one of the key indicators for judging the severity of the internal situation. A slow pressure accumulation may stem from normal thermal expansion and contraction or slight gas production, while a sharp, soaring rate of pressure increase strongly indicates that violent electrolyte decomposition, combustion, or other intense chemical reactions are occurring. This submodule compares the real-time calculated pressure values ​​and rates of pressure change with preset safe pressure ranges. The safe pressure ranges are determined based on the battery pack's structural design strength, sealing performance, and safety margins. When the pressure value or rate of pressure change approaches or exceeds these safety thresholds, it indicates that the battery pack structure is facing a risk.

[0067] The pressure relief strategy adjustment submodule is the core of pressure management decision-making. It receives pressure risk assessment results from the pressure trend analysis submodule, as well as thermal runaway suppression plans from the emergency suppression module. This is because suppression actions themselves can significantly affect the internal pressure status. This submodule needs to integrate information from both sources to formulate the optimal pressure relief strategy. Its decision-making includes determining the optimal timing for opening the pressure relief valves. Opening too early may result in premature loss of potentially effective inhibitors or the introduction of oxygen; opening too late may expose the enclosure to excessive pressure risk. The decision also requires determining the opening extent of the pressure relief valves (e.g., percentage of opening area or opening duration), whether to perform a gentle, gradual release or a rapid, forceful venting. This depends on the current pressure hazard level and the inhibitor deployment plan. Finally, this submodule generates a pressure control instruction, clearly specifying which pressure relief valves(s) should operate at what time and in what manner (opening extent, duration) to ensure that the internal pressure remains under control, creating a safe internal environment for emergency suppression.

[0068] The core thermal balance model used in the dose calculation submodule of the emergency suppression module can be simplified as follows:

[0069]

[0070] All temperature measurements are in Kelvin. and They are different and the units are also different. This indicates the calculated mass of the required suppression medium (in kilograms). This indicates the total number of modules or components that need to be cooled. Indicates the first The mass (in kilograms) of a module or component that requires cooling. Indicates the first The specific heat capacity of a module or material (in joules per kilogram per Kelvin). This indicates the target safe temperature (in Kelvin) to which these components are expected to be cooled. Indicates the first The current real-time temperature of each component (in Kelvin). This represents the estimated additional heat (in joules) that the ongoing thermal runaway reaction may release in the subsequent time. This indicates the latent heat of vaporization of the selected suppressing medium (in joules per kilogram). This indicates the specific heat capacity of the selected suppressing medium in the gaseous state (unit: joules per kilogram per Kelvin). This indicates the final temperature reached by the inhibitor gas (in Kelvin). This indicates the initial temperature of the inhibitor in the storage tank (in Kelvin).

[0071] This formula quantifies the relationship between the total heat to be transferred and the heat that the selected medium can absorb, serving as the theoretical basis for the precise delivery achieved by the dosage calculation submodule. The entire implementation embodies a complete closed loop from suppression decision-making to pressure coordination control, tightly integrating predictive algorithms with physical actuators.

[0072] Example 4: See Figure 4 The multi-mode linkage module, acting as the system's command center, begins its operation with the data integration submodule. This submodule simultaneously receives asynchronous data streams from multiple upstream modules: battery thermal runaway early warning signals generated by the thermal signal monitoring module, thermal runaway prediction results generated by the thermal runaway prediction module, thermal runaway suppression plans formulated by the emergency suppression module, and pressure control commands issued by the pressure balance module. These data streams not only differ in content but also in their timestamps and update frequencies. The core task of the data integration submodule is to align and fuse these multi-source heterogeneous data along the time dimension. It maintains a system global clock, assigns a unified timestamp to each input data packet, and uses interpolation or buffering mechanisms to synchronize data arriving at different times to the same decision-making point in time, forming a unified situational awareness encompassing four dimensions: heat, runaway, suppression, and pressure. This time-aligned multi-source data set provides a consistent and comparable information foundation for subsequent priority calculations.

[0073] The response priority calculation submodule performs in-depth analysis of the integrated multi-source data to assess the overall risk situation and determine the order of response actions. This submodule's assessment is based on a multi-dimensional weighted scoring algorithm. The algorithm first divides the battery pack's interior into virtual regions, extracting corresponding characteristic parameters for each region from the integrated data: the thermal runaway risk level and temperature change acceleration from the thermal signal dimension; the estimated time and heat flux intensity of the region being affected by heat waves from the diffusion prediction dimension; and the real-time pressure value and pressure rise rate from the pressure dimension. Each parameter is assigned a weight coefficient, which is preset based on its impact on safety. The submodule calculates a comprehensive "risk score" for each region; regions with higher scores indicate greater danger and a more urgent need for priority response. Subsequently, all pending response actions within the system are evaluated and prioritized based on factors such as the risk areas they affect, expected effects, and execution time. Finally, the submodule outputs a global action priority list, clearly defining which actions should be executed first, which can be executed in parallel, and which need to be postponed in complex scenarios.

[0074] The linkage instruction generation submodule is the final stage in converting decisions into actions. It receives the action priority list output by the response priority calculation submodule and converts it into a series of coordinated operation instructions that can be directly recognized and executed by the underlying actuators. These instructions not only contain the action content itself but also specify the execution sequence, duration, and intensity level of the actions in greater detail. For example, the instructions might explicitly require: at time T0, inhibitor nozzle No. 1 in zone A should open first, releasing perfluorohexanone at 60% flow rate for 200 milliseconds; at time T0+50 milliseconds, nozzle No. 2 in zone B downstream of the heat flow should activate for interception; simultaneously, at time T0+20 milliseconds, pressure relief valve No. 7 at the top of the package should open to 30% for preventative pressure relief. All these instructions are encoded into control signals and sent in parallel and synchronously to the actuators (solenoid valves, pumps, pressure relief valve actuators, etc.) distributed throughout the system via the system control bus, thus forming a system-level emergency response instruction that is precisely coordinated in both time and space.

[0075] The signal verification submodule within the thermal signal monitoring module serves as a crucial guarantee of data reliability. This submodule activates after the temperature signal acquisition submodule obtains initial readings from multiple sensors within the same area. Its design aims to identify and eliminate unreliable data caused by sensor malfunctions, transient interference, or localized measurement anomalies. The signal verification submodule employs a consistency check method, calculating the statistical distribution of all sensor readings within the same area group, such as calculating their mean and standard deviation. It sets an allowable deviation threshold; when a sensor reading deviates from the group's average beyond this threshold, the reading is marked as "suspicious" or "abnormal." For marked abnormal readings, the submodule removes them from the current dataset, excluding them from the overall temperature calculation for that area. Subsequently, it recalculates the aggregated temperature value and trend for the area using readings from other functioning sensors within the region. This mechanism effectively prevents erroneous warnings from a single sensor malfunction or false alarm, significantly improving the robustness and accuracy of thermal signal monitoring.

[0076] To illustrate the evaluation process of the response priority calculation submodule more specifically, a simplified regional risk assessment table is shown below. Table 1 simulates a segment of the system's risk assessment of four different regions within the battery pack at a certain moment.

[0077] Table 1: Regional Risk Assessment and Priority Ranking.

[0078]

[0079] Referring to Table 1, this table reflects the system's snapshot assessment of the four zones at a specific moment. It can be seen that Zone_05, due to its high thermal risk level and shortest predicted heating time, has the highest overall risk score despite not having the highest pressure rise rate, and is therefore listed as the highest priority zone. Zone_12, although having a "medium" thermal risk level, has an exceptionally high pressure rise rate that contributes a significant risk weight, placing it second in priority. Zone_03, while having a high thermal risk level, has relatively gradual pressure changes and a later onset of thermal impact, thus its priority is lower than the previous two. Zone_08 has low scores across all indicators, therefore its priority is the lowest. This process demonstrates that the system does not rely on a single indicator but rather performs multi-factor weighted fusion decision-making, thereby arriving at a more comprehensive and reasonable order of action.

[0080] It demonstrates how the system effectively integrates and deeply processes information from different sources and of different natures, generates globally optimal response strategies through rigorous evaluation logic, and ensures the reliability of decision-making basis through signal verification mechanisms, ultimately achieving a precise, orderly, and efficient multi-mode linkage emergency response.

[0081] Example 5: The real-time correction submodule in the thermal diffusion prediction module plays a crucial role in dynamically calibrating the thermal diffusion prediction. This submodule continuously receives the latest temperature data stream from the thermal signal monitoring module, reflecting the actual evolution of the thermal field within the battery pack. Its core function is to continuously compare these real-time observed temperature distributions and trends with the heat propagation paths and diffusion rates predicted earlier by the thermal diffusion prediction module based on initial conditions. This comparison is not a simple numerical comparison, but a process of pattern matching and difference analysis. The submodule's built-in algorithm analyzes the deviation between the actual heat front's movement speed and the predicted speed, monitoring whether unexpected propagation directions of heat have occurred (e.g., due to a difference in the thermal conductivity of a structural component compared to the model's preset value, heat may unexpectedly accelerate its transfer to a certain area). When a significant and persistent systematic deviation between the predicted and measured values ​​is identified, the real-time correction submodule initiates a correction procedure. It does not completely discard the initial model but treats the real-time data as a feedback signal to dynamically adjust key parameters in the prediction model. These parameters may include the equivalent thermal conductivity along a specific path, the estimated contact thermal resistance between modules, or the weighting of convective heat transfer effects. In this way, the submodules can dynamically adjust the predictions of heat transfer rate and diffusion direction online, continuously approximating the actual physical process and narrowing the gap between prediction and reality. This dynamic correction mechanism enables the system to maintain the reliability of predictions when facing complex, variable, and potentially uncertain real-world thermal runaway scenarios, providing more accurate time windows and situational information for subsequent emergency response decisions.

[0082] The feedback learning submodule within the multi-mode linkage module focuses on long-term system performance improvement and knowledge accumulation. This submodule begins operation after each system triggers and completes its emergency response process. Its primary task is to comprehensively collect and record relevant data for the event, forming a complete "response case" record. This execution effect data is extensive, including but not limited to: the complete temperature time series from the initial warning characteristics recorded by the thermal signal monitoring module to the evolution of thermal runaway; the initial predictions output by the thermal diffusion prediction module and their subsequent correction records; details of the suppression plan actually executed by the emergency suppression module, including medium type, dosage, delivery point, and timing; the execution status of pressure control commands from the pressure balancing module; and the final system-level response commands and their execution results. More importantly, it also records key outcome indicators, such as whether thermal runaway was successfully suppressed, the total time required for suppression, whether the range of thermal diffusion was effectively controlled, whether the pressure smoothly returned to a safe range, and whether any secondary problems occurred.

[0083] After collecting this detailed case data, the core learning process of the feedback learning submodule unfolds. It integrates data mining and machine learning algorithms to perform in-depth analysis of this multi-dimensional time-series data. This process aims to discover new patterns, validate the effectiveness of existing strategies, and identify areas for improvement. For example, the algorithm might analyze and discover that for a specific pattern of initial temperature rise, the current standard suppression scheme's response threshold may be too conservative, and intervention at an earlier stage could be attempted. Or, it might discover that, under a certain battery pack layout, a combination of a pressure relief valve and a specific nozzle consistently achieves faster pressure stabilization than other combinations.

[0084] Based on the insights derived from these analyses, the feedback learning submodule performs knowledge updates. It adds feature data of the current event to the historical thermal runaway case feature library relied upon by the thermal signal monitoring module, especially features from successfully suppressed or typical cases, thereby enriching and diversifying the sample library for future pattern matching. Simultaneously, it fine-tunes and optimizes the judgment rule set relied upon by the thermal risk assessment submodule, such as adjusting the risk level weights corresponding to different temperature change acceleration ranges, or introducing newly identified early warning features as auxiliary criteria. The optimization of response strategies is more comprehensive; it may suggest adjusting the weight allocation of the priority calculation submodule in the multi-mode linkage module, or providing corrected parameters for the dosage calculation model of the emergency suppression module, or even recommending new or optimized inhibitor delivery strategy templates for specific scenarios.

[0085] This embodies a closed-loop process from execution to learning, and then from learning to optimization. The real-time correction submodule ensures the system's short-term adaptability in responding to a single event, while the feedback learning submodule drives the system's long-term evolution through the accumulation of multiple events. These two modules complement each other, enabling the system to move beyond relying solely on initial, fixed models and rules, becoming an intelligent system capable of continuously learning from real-world operating environments, accumulating experience, and constantly improving its early warning accuracy and emergency response effectiveness. This self-optimization capability is crucial for addressing complex and diverse lithium-ion battery thermal runaway scenarios.

[0086] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0087] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A multi-level joint prevention and emergency response system for thermal runaway of lithium-ion batteries, characterized in that, The system includes: The thermal signal monitoring module acquires multi-dimensional temperature sensor data inside the battery pack, identifies abnormal temperature signals and temperature change trends, and generates a battery thermal runaway early warning signal. The thermal runaway prediction module calculates the heat propagation path and diffusion rate based on the battery thermal runaway early warning signal and the battery module structural parameters, and generates thermal runaway prediction results. The emergency suppression module matches the suppression medium and calculates the inhibitor dosage strategy based on the thermal diffusion prediction results, and generates a thermal runaway suppression scheme. The pressure balance module acquires pressure sensor data inside the battery pack, analyzes the rate of pressure change and pressure distribution, adjusts the pressure relief valve action strategy according to the thermal runaway suppression scheme, and generates pressure control commands. Based on the pressure control command, the multi-mode linkage module integrates thermal signal data, diffusion prediction results, and suppression schemes to generate system-level emergency response commands.

2. The multi-level joint prevention and emergency response system for thermal runaway of lithium-ion batteries according to claim 1, characterized in that, The battery thermal runaway early warning signal includes an abnormal temperature area identifier, a temperature change acceleration value, and a thermal runaway risk level. The thermal diffusion prediction results include a predicted heat propagation direction, a heat transfer rate between modules, and an estimated critical diffusion time. The thermal runaway suppression scheme includes the type of suppression medium, the dosage of the medium, and the distribution of the medium's placement location. The pressure control command includes the timing of the pressure relief valve opening, the duration of pressure relief, and the pressure release intensity classification. The system-level emergency response command includes multi-level response trigger conditions, system action priority ranking, and linkage control timing.

3. The multi-level joint prevention and emergency response system for lithium-ion battery thermal runaway according to claim 2, characterized in that, The thermal signal monitoring module includes: The temperature signal acquisition submodule acquires data from multiple temperature sensors arranged inside the battery pack, aggregates temperature readings by region, and records the temperature value for each acquisition cycle. The temperature change analysis submodule calculates the temperature change difference between adjacent time points based on the temperature value, and determines whether the temperature change exceeds the safety threshold by combining the normal operating temperature range of the battery. The thermal risk assessment submodule compares the temperature change difference with the temperature characteristics of historical thermal runaway cases and outputs a battery thermal runaway warning signal.

4. The multi-level joint prevention and emergency response system for thermal runaway of lithium-ion batteries according to claim 3, characterized in that, The thermal diffusion prediction module includes: Based on the battery thermal runaway early warning signal, the module thermal conduction analysis submodule calls the thermal conduction coefficient between battery modules and the structural gap parameters to calculate the heat transfer rate between modules. The diffusion path simulation submodule simulates possible heat propagation paths based on the heat transfer rate between modules and the battery pack spatial layout data. The thermal impact prediction submodule estimates the time it takes for heat to reach critical components based on the possible propagation paths and generates thermal diffusion prediction results.

5. The multi-level joint prevention and emergency response system for lithium-ion battery thermal runaway according to claim 4, characterized in that, The emergency suppression module includes: Based on the thermal diffusion prediction results, the suppression medium matching submodule queries the medium property database and selects a suppression medium that matches the thermal runaway type. The dosage calculation submodule calculates the required amount of suppression medium to be administered based on the heat transfer rate and the range of influence in the heat diffusion prediction results. The deployment strategy generation submodule, in conjunction with the internal structure of the battery pack, determines the release points and sequence of the suppression medium, and generates a thermal runaway suppression scheme.

6. The multi-level joint prevention and emergency response system for thermal runaway of lithium-ion batteries according to claim 5, characterized in that, The pressure balancing module includes: The pressure data acquisition submodule acquires real-time data from multiple pressure sensors inside the battery pack and records the pressure values ​​and change times by region. The pressure trend analysis submodule calculates the rate of pressure rise per unit time based on the real-time data and compares it with the safe pressure range. The pressure relief strategy adjustment submodule determines the triggering timing and opening range of the pressure relief valve based on the pressure rise rate and thermal runaway suppression scheme, and generates pressure control commands.

7. The multi-level joint prevention and emergency response system for lithium-ion battery thermal runaway according to claim 6, characterized in that, The multi-mode linkage module includes: The data integration submodule receives the battery thermal runaway early warning signal, thermal diffusion prediction results, thermal runaway suppression scheme and pressure control command, and aligns the multi-source data according to the time series. The response priority calculation submodule assesses the thermal risk level and pressure hazard level of different regions based on the multi-source data, and determines the priority order of response actions; The linkage instruction generation submodule generates coordinated operation instructions, including suppression, pressure relief, and alarm, according to the aforementioned priority order, and outputs system-level emergency response instructions.

8. The multi-level joint prevention and emergency response system for thermal runaway of lithium-ion batteries according to claim 7, characterized in that, The thermal signal monitoring module also includes a signal verification submodule, which acquires temperature data from multiple sensors in the same area, compares the consistency of the readings of each sensor, eliminates abnormal readings, and recalculates the temperature change trend.

9. The multi-level joint prevention and emergency response system for thermal runaway of lithium-ion batteries according to claim 8, characterized in that, The heat diffusion prediction module also includes a real-time correction submodule, which receives the latest temperature data, compares it with the predicted diffusion path, and dynamically adjusts the predicted heat transfer rate and diffusion direction.

10. The multi-level joint prevention and emergency response system for thermal runaway of lithium-ion batteries according to claim 9, characterized in that, The multi-mode linkage module also includes a feedback learning sub-module, which records the execution effect data of each emergency response, updates the historical case feature library, and optimizes the thermal risk judgment rules and response strategies.

Images