Battery system multi-valve coordinated explosive pressure release method and system

Abstract

This invention relates to the field of battery safety assessment technology, and particularly to a method and system for multi-valve linkage pressure release in battery systems. The method involves analyzing multi-dimensional time-series state characteristic data of the battery system based on battery system operational status monitoring data; performing thermal runaway propagation risk characteristic analysis based on the multi-dimensional time-series state characteristic data to generate thermal runaway propagation risk characteristic data; conducting multi-valve linkage control topology analysis of the battery system based on the battery casing pressure relief valve structure configuration data and the thermal runaway propagation risk characteristic data to generate multi-valve linkage control topology data; and designing a multi-valve linkage pressure release control strategy based on the multi-valve linkage control topology data to generate an intelligent control strategy for multi-valve linkage pressure release in battery systems. This invention achieves safe and efficient multi-valve linkage pressure release in battery systems.

Description

Technical Field

[0001] This invention relates to the field of battery safety assessment technology, and in particular to a method and system for releasing burst pressure in a battery system with multiple valve linkages. Background Technology

[0002] With the rapid development of the new energy industry, power battery systems are widely used in new energy vehicles, large-scale energy storage, and other fields. Their safety has become a key bottleneck restricting the high-quality development of the industry. Power battery systems are highly susceptible to thermal runaway during charge-discharge cycles, long-term service, or extreme conditions. During thermal runaway, a large amount of high-temperature, high-pressure gas is rapidly generated, accompanied by a dramatic pressure increase. If the internal pressure cannot be released in a timely and effective manner, it can lead to accidents such as battery casing rupture, fire, or even explosion. However, existing battery system pressure release methods mostly employ a single pressure relief valve structure or a simple multi-valve parallel opening mode, which is insufficient to cope with the instantaneous high-pressure impact during large-scale battery system thermal runaway. Furthermore, this can easily lead to pressure relief disorder, causing local pressure accumulation or uneven energy release, which in turn exacerbates structural damage to the battery system. Monitoring of battery system thermal runaway is often limited to a single physical quantity, making it difficult to accurately capture the propagation patterns and risk accumulation characteristics of thermal runaway. This results in a lack of scientific and accurate decision-making basis for multi-valve linkage control, hindering the realization of intelligent and refined control of pressure release. Summary of the Invention

[0003] Based on this, the present invention provides a method and system for releasing burst pressure in a battery system with multiple valves to solve at least one of the above-mentioned technical problems.

[0004] To achieve the above objectives, a multi-valve linkage method for releasing burst pressure in a battery system includes the following steps: Step S1: Utilize a multi-type sensor array deployed inside the battery module and casing to monitor and process the battery system's operating status, generating battery system operating status monitoring data; based on the battery system operating status monitoring data, perform multi-dimensional time-series state characteristic analysis of the battery system, generating multi-dimensional time-series state characteristic data of the battery system. Step S2: Analyze the thermal runaway propagation risk characteristics of the battery system based on the multi-dimensional time-series state characteristic data of the battery system, and generate thermal runaway propagation risk characteristic data of the battery system; Step S3: Obtain the structural configuration data of the battery housing pressure relief valve; perform multi-valve linkage control topology analysis on the multi-dimensional time-series state characteristic data of the battery system using the structural configuration data of the battery housing pressure relief valve and the thermal runaway propagation risk characteristic data of the battery system, and generate multi-valve linkage control topology data of the battery system. Step S4: Design the burst pressure release control strategy for the multi-valve linkage of the battery system based on the multi-valve linkage control topology data of the battery system, and generate the intelligent control strategy for burst pressure release of the multi-valve linkage of the battery system.

[0005] Furthermore, step S1 includes the following steps: Step S11: Use a multi-type sensor array deployed inside the battery module and casing to monitor and process the battery system's operating status, and generate battery system operating status monitoring data; Step S12: Perform multi-source response analysis of battery system operating status based on battery system operating status monitoring data to generate multi-source response data of battery system operating status; Step S13: Perform multi-physics coupling deconstruction processing on the battery system based on the multi-source response data of the battery system operating status to generate multi-physics coupling deconstruction data of the battery system; Step S14: Perform battery system response decoupling characteristic analysis based on the multi-physics coupling deconstruction data of the battery system to generate battery system response decoupling characteristic data; Step S15: Perform spatial energy and matter migration mapping feature analysis using battery system response decoupling feature data to generate battery system energy-matter migration mapping feature data; Step S16: Based on the battery system response decoupling feature data and the battery system energy-mass migration mapping feature data, perform adaptive dimensionality reduction and resampling processing of the battery system response migration sensitive region to generate dimensionality reduction response decoupling feature data of the battery system. Step S17: Perform multi-dimensional time-series state feature analysis on the battery system distribution dimensionality reduction of the battery system response decoupling feature data and the battery system energy-mass migration mapping feature data through the battery system dimensionality reduction response decoupling feature data, and generate multi-dimensional time-series state feature data of the battery system.

[0006] Furthermore, the multi-source response data of the battery system operating status mentioned in step S12 includes temperature dynamic response data, gas concentration change response data, internal pressure transient fluctuation response data, voltage and current disturbance response data, and casing strain response data.

[0007] Furthermore, step S14 includes the following steps: Step S141: Perform multi-physics decoupling response gradient distribution analysis on the multi-physics coupling deconstruction data of the battery system to generate multi-physics decoupling response gradient distribution data; Step S142: Perform dominant feature analysis of local response of multiphysics decoupling based on the gradient distribution data of multiphysics decoupling response, and generate dominant feature data of local response of multiphysics decoupling response; Step S143: Perform correlation analysis on the multi-physics decoupling based on the multi-physics coupling deconstruction data of the battery system, and generate multi-physics decoupling correlation data; Step S144: Perform battery system response decoupling feature analysis using multi-physics decoupling local response dominant feature data and multi-physics decoupling correlation data to generate battery system response decoupling feature data.

[0008] Furthermore, step S2 includes the following steps: Step S21: Perform multi-layer heterogeneous characteristic analysis of the battery system based on the multi-dimensional time-series state characteristic data of the battery system, generate multi-layer heterogeneous characteristic data of the battery system, and perform multi-layer heterogeneous structural node modeling processing through the multi-layer heterogeneous characteristic data of the battery system to generate multi-layer heterogeneous structural node data. Step S22: Analyze the multi-mechanism propagation relationship of the battery system based on the multi-layer heterogeneous structure node data, generate multi-mechanism propagation relationship data of the battery system, and analyze the multi-mechanism propagation driving factors through the multi-mechanism propagation relationship data of the battery system to generate multi-mechanism propagation driving factor data. Step S23: Analyze the propagation path distribution characteristics based on the multi-layer heterogeneous structure node data and the multi-mechanism propagation driving factor data to generate propagation path distribution characteristic data; Step S24: Analyze the risk clustering characteristics of the transmission path based on the distribution characteristic data of the transmission path, and generate risk clustering characteristic data of the transmission path; Step S25: Analyze the propagation risk characteristics of the battery system through the propagation path risk clustering feature data, and generate the thermal runaway propagation risk feature data of the battery system.

[0009] Furthermore, step S3 includes the following steps: Step S31: Obtain the structural configuration data of the battery housing pressure relief valve; Step S32: Based on the battery casing pressure relief valve structure configuration data and battery system thermal runaway propagation risk characteristic data, perform thermal runaway risk and valve action mapping relationship analysis to generate thermal runaway risk-valve action mapping relationship data; Step S33: Based on the thermal runaway risk-valve action mapping relationship data, perform multi-valve linkage candidate combination analysis for thermal runaway risk to generate multi-valve linkage candidate combination data; perform constraint optimization solution on the multi-valve linkage candidate combination data to generate multi-valve linkage combination optimization data; Step S34: Based on the multi-valve linkage combination optimization data, perform multi-valve linkage control topology analysis on the multi-dimensional time-series state characteristic data of the battery system to generate multi-valve linkage control topology data of the battery system.

[0010] Furthermore, step S32 includes the following steps: Step S321: Based on the battery housing pressure relief valve structure configuration data, perform pressure relief valve execution node modeling to generate battery system pressure relief valve execution node data; Step S322: Perform a function correlation feature analysis on the pressure relief valve execution node data of the battery system to generate function correlation feature data of the pressure relief valve execution node; Step S323: Perform relational abstraction processing on the thermal runaway propagation risk characteristic data of the battery system to generate abstract characteristic data of thermal runaway propagation risk; Step S324: Map the function-related feature data of the pressure relief valve to the abstract feature data of thermal runaway propagation risk to perform thermal runaway risk and valve function mapping relationship analysis, and generate thermal runaway risk-valve function mapping relationship data.

[0011] Furthermore, step S4 includes the following steps: Step S41: Analyze the factors affecting the burst pressure of the multi-valve linkage control topology based on the multi-valve linkage control topology data of the battery system, and generate multi-valve linkage burst pressure influencing factor data; Step S42: Perform heterogeneous combination feature analysis of multi-valve linkage based on the multi-valve linkage control topology data of the battery system, and generate heterogeneous combination feature data of multi-valve linkage. Step S43: Perform nonlinear time-series trend analysis of the influencing factors of multi-valve linkage explosion pressure using multi-valve linkage explosion pressure influencing factor data and multi-valve linkage heterogeneous combination characteristic data, and generate multi-valve linkage explosion pressure influencing factor trend data. Step S44: Perform dynamic prediction processing of explosion pressure evolution based on the trend data of multi-valve linkage explosion pressure influencing factors to generate dynamic prediction data of explosion pressure evolution; Step S45: Design the explosion pressure release control strategy for the multi-valve linkage of the battery system using the multi-valve linkage control topology data and the dynamic prediction data of explosion pressure evolution, and generate the intelligent control strategy for explosion pressure release of the multi-valve linkage of the battery system.

[0012] Furthermore, step S45 includes the following steps: Step S451: Perform action space analysis of multi-valve collaborative control strategy based on the multi-valve linkage control topology data of the battery system, and generate action space data of multi-valve collaborative control strategy; Step S452: Based on the action space data of the multi-valve collaborative control strategy, perform an action space explosion pressure release benefit analysis on the dynamic prediction data of explosion pressure evolution, and generate action space explosion pressure release benefit data; Step S453: Analyze the optimization requirements of the battery system's explosion pressure evolution using dynamic prediction data of explosion pressure evolution, and generate optimization requirement data for the battery system's explosion pressure evolution. Step S454: Design a multi-valve linkage intelligent control strategy for the release of burst pressure in the battery system based on the burst pressure release benefit data of the action space and the target data of the burst pressure evolution optimization requirements of the battery system.

[0013] This specification provides a multi-valve linkage burst pressure release system for a battery system, used to execute the multi-valve linkage burst pressure release method for a battery system as described above. The multi-valve linkage burst pressure release system for a battery system includes: The battery system operation status monitoring and analysis module is used to monitor and process the battery system operation status using a multi-type sensor array deployed inside the battery module and casing, and generate battery system operation status monitoring data; based on the battery system operation status monitoring data, it performs multi-dimensional time-series state characteristic analysis of the battery system, and generates multi-dimensional time-series state characteristic data of the battery system. The thermal runaway propagation risk characteristic analysis module is used to perform thermal runaway propagation risk characteristic analysis on battery system based on multi-dimensional time-series state characteristic data of battery system, and generate thermal runaway propagation risk characteristic data of battery system; The battery system multi-valve linkage control topology analysis module is used to acquire the structural configuration data of the battery housing pressure relief valve; and to perform multi-valve linkage control topology analysis on the multi-dimensional time-series state characteristic data of the battery system using the structural configuration data of the battery housing pressure relief valve and the thermal runaway propagation risk characteristic data of the battery system, thereby generating multi-valve linkage control topology data of the battery system. The intelligent control module for multi-valve linkage pressure release in battery systems is used to design pressure release control strategies for multi-valve linkage in battery systems based on the multi-valve linkage control topology data of the battery system, and to generate intelligent control strategies for multi-valve linkage pressure release in battery systems.

[0014] The beneficial effects of this application are as follows: This invention acquires real and comprehensive basic monitoring data through a multi-type sensor array, and then clarifies the multi-dimensional response characteristics of the battery system under different operating states through multi-source response analysis, making the monitoring data more targeted. Subsequently, through a series of processes such as multi-physics coupling deconstruction, response decoupling, spatial energy and matter migration mapping analysis, and adaptive dimensionality reduction and resampling of response migration sensitive areas, redundant data is effectively eliminated, while accurately capturing the temporal variation characteristics and response sensitive areas of the battery system. This avoids subsequent analysis biases caused by complex data and unclear characteristics, ensuring that the generated multi-dimensional temporal state characteristic data can truly reflect the real-time operating state of the battery system, laying a solid technical foundation for subsequent thermal runaway propagation risk analysis, and significantly improving the accuracy and reliability of the entire explosion pressure release control method. In addition, during the response decoupling feature analysis process, gradient distribution analysis, local response dominant feature analysis, and correlation analysis further optimize the extraction accuracy of response decoupling features, ensuring that the decoupling data can accurately reflect the independent effects and interrelationships of each physical field. Through multi-layer heterogeneous characteristic analysis and structural node modeling, the structural characteristics and node relationships of the battery system were clearly identified, providing a clear analytical framework for propagation relationship analysis. Subsequently, through multi-mechanism propagation relationship analysis, driving factor analysis, propagation path distribution, and risk aggregation characteristic analysis, the core laws of thermal runaway propagation were gradually deconstructed, the propagation path and risk aggregation area were accurately located, and the driving factors of thermal runaway propagation were clarified. This allows for the early prediction of the propagation trend and risk level of thermal runaway, avoiding the problem of lag in multi-valve linkage control caused by untimely or inaccurate prediction of thermal runaway propagation. This provides accurate decision-making basis for subsequent multi-valve linkage control topology analysis and burst pressure release control strategy design, ensuring that subsequent multi-valve linkage control can specifically address the risk of thermal runaway propagation, improving the timeliness and effectiveness of burst pressure release, further ensuring the operational safety of the battery system, and making up for the lack of scientific decision support in multi-valve linkage control. Obtaining relevant data on the structural configuration of pressure relief valves provided a foundation for subsequent topology analysis. Subsequently, through modeling the execution nodes of the pressure relief valves and analyzing their functional correlation characteristics, the execution characteristics and interrelationships of each pressure relief valve were clearly identified. Simultaneously, the characteristics of thermal runaway propagation risk were abstracted, achieving a precise mapping between the function of the pressure relief valves and the risk of thermal runaway, avoiding the disconnect between valve function and risk control. Based on this, through multi-valve linkage candidate combination analysis and constraint optimization, the optimal multi-valve linkage combination scheme was selected, eliminating unreasonable linkage combinations. This ensures that multi-valve linkage control can efficiently adapt to the risk of thermal runaway propagation, clarifies the control logic and correlation of multi-valve linkage, overcomes the limitations of multi-valve parallel control, makes multi-valve linkage control more targeted and scientific, and further improves the overall effectiveness of pressure release control.By analyzing the factors influencing burst pressure and the characteristics of heterogeneous combinations in a multi-valve linkage control topology, the core factors affecting burst pressure release and the combined characteristics of multi-valve linkage were accurately identified, providing comprehensive support for strategy design. Subsequently, through nonlinear time-series trend analysis of burst pressure influencing factors and dynamic prediction of burst pressure evolution, the changing trend and evolution law of burst pressure can be predicted in advance, avoiding the problem of control strategy lagging behind burst pressure changes. In the strategy design process, through action space analysis, burst pressure release benefit analysis, and optimization requirement target analysis of the multi-valve collaborative control strategy, the optimal control action scheme is selected to ensure that the control strategy can take into account both burst pressure release benefits and battery system safety. The generated multi-valve linkage burst pressure release intelligent control strategy can flexibly adjust the control mode of multiple valves according to the dynamic evolution of burst pressure and the risk of thermal runaway propagation, achieving precise and efficient release of burst pressure, effectively avoiding problems such as untimely pressure release and pressure release disorder caused by unreasonable control strategies.

[0015] Therefore, the multi-valve linkage pressure release method for battery systems of the present invention, through precise monitoring of the battery system's operating status and dynamic analysis of the risk of thermal runaway propagation, adjusts the opening timing, sequence, and degree of multiple valves according to the real-time status of the battery system. This effectively solves the problems of low pressure relief efficiency and limited range of a single pressure relief valve, as well as the problems of pressure relief disorder, local pressure accumulation, and aggravated structural damage caused by simple parallel opening of multiple valves. It significantly improves the efficiency and balance of pressure release, while taking into account the stability and safety of the battery system during pressure relief, and can effectively cope with the instantaneous high-pressure impact during large-scale battery system thermal runaway. By collecting multi-dimensional data such as temperature, gas concentration, internal pressure, and shell strain through a multi-type sensor array, combined with multi-physics field coupling deconstruction and decoupling analysis, it achieves precise capture of the thermal runaway propagation law and risk accumulation characteristics, providing a scientific and accurate decision-making basis for multi-valve linkage control. This solves the shortcomings of existing multi-valve linkage control, which lacks effective decision support and cannot achieve intelligent and refined control, realizing intelligent and refined management of pressure release and further improving the safety protection level of the battery system. Attached Figure Description

[0016] Figure 1 This is a schematic diagram of the steps of a multi-valve linkage explosion pressure release method for a battery system according to the present invention; Figure 2 for Figure 1 A detailed flowchart illustrating the implementation steps of step S3. The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0017] The technical method of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0018] Furthermore, the accompanying drawings are merely illustrative of the invention and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and therefore repeated descriptions of them will be omitted. Some block diagrams shown in the drawings are functional entities and do not necessarily correspond to physically or logically independent entities. Functional entities may be implemented in software, or in one or more hardware modules or integrated circuits, or in different network and / or processor methods and / or microcontroller methods. The term "and / or" as used herein includes any and all combinations of one or more of the associated items listed.

[0019] To achieve the above objectives, please refer to Figures 1 to 2 This invention provides a multi-valve linkage method and system for releasing burst pressure in a battery system. In the embodiments of this invention, please refer to... Figure 1 The diagram shown is a flowchart illustrating the steps of a multi-valve linkage explosion pressure release method for a battery system according to the present invention. The multi-valve linkage explosion pressure release method for a battery system includes the following steps: Step S1: Utilize a multi-type sensor array deployed inside the battery module and casing to monitor and process the battery system's operating status, generating battery system operating status monitoring data; based on the battery system operating status monitoring data, perform multi-dimensional time-series state characteristic analysis of the battery system, generating multi-dimensional time-series state characteristic data of the battery system. In this embodiment of the invention, a multi-type sensor array deployed inside the battery module and casing is used to monitor the operating status of the battery system. A multi-type sensor array, consisting of temperature sensors, gas sensors, internal pressure sensors, voltage and current sensors, and strain sensors, is deployed on the surface of each cell in the battery module, in the gaps between modules, and in different key areas inside the battery casing. The sensor array continuously collects various physical parameters during the operation of the battery system at fixed intervals, comprehensively capturing the state changes of the battery system at different operating stages. All collected parameters are integrated to form battery system operating status monitoring data, ensuring that the monitoring data covers all core parts and key physical quantities of the battery system, providing complete and accurate basic data for subsequent feature analysis. Based on the monitoring data, a multi-dimensional time-series state characteristic analysis of the battery system was conducted. First, multi-source response analysis was performed on the monitoring data to distinguish the response characteristics of different physical parameters. Then, the temperature field, pressure field, electric field, and strain field inside the battery system were coupled and deconstructed to dissect the interaction relationships between the various physical fields. Subsequently, the deconstructed features were decoupled and analyzed to extract the independent characteristic parameters of each physical field. At the same time, the migration laws of spatial energy and matter were analyzed. Adaptive dimensionality reduction and resampling were used to remove redundant data and retain core features. Finally, the time-series variation laws of each parameter were analyzed by combining the dimensionality-reduced feature data, and the multi-dimensional time-series state characteristic data of the battery system was integrated to clearly present the dynamic changes and core characteristics of the battery system's operating state. This provides accurate data support for subsequent thermal runaway propagation risk analysis and ensures the accuracy of the entire pressure release method.

[0020] Step S2: Analyze the thermal runaway propagation risk characteristics of the battery system based on the multi-dimensional time-series state characteristic data of the battery system, and generate thermal runaway propagation risk characteristic data of the battery system; In this embodiment of the invention, a risk characteristic analysis of thermal runaway propagation in a battery system is conducted based on multi-dimensional temporal state characteristic data. The core analytical basis is the temporal variation characteristics of each physical field and the laws governing energy and matter migration within the multi-dimensional temporal state characteristic data. First, the battery system undergoes multi-layer heterogeneous characteristic analysis, dividing it into different levels based on cells, modules, and casings. Structural features, temporal variation characteristics, and energy migration characteristics of each level are extracted. Based on the features of each level, structural nodes are modeled to clarify the relationships and distribution of nodes at each level. Subsequently, the multi-mechanism propagation relationship of thermal runaway in the battery system is analyzed based on the structural node data. The propagation paths and interactions between nodes under the three core propagation mechanisms of heat conduction, gas diffusion, and electrical conduction are dissected. Simultaneously, the core driving factors of each propagation mechanism are analyzed to identify the key factors driving thermal runaway propagation. Based on this, the distribution characteristics of thermal runaway propagation paths are analyzed, tracking the direction, coverage, and propagation intensity of each path. Then, risk aggregation characteristic analysis is performed on each propagation path to locate risk aggregation nodes and high-risk aggregation areas, clarifying the degree of risk aggregation and its evolutionary laws. By integrating all analysis results, the core risk characteristics of thermal runaway propagation are extracted, the classification criteria, risk distribution and evolution trend of different risk levels are clarified, and risk characteristic data of thermal runaway propagation in battery systems are generated. This provides accurate risk basis for subsequent multi-valve linkage control topology analysis, ensuring that multi-valve linkage control can address thermal runaway propagation risks in a targeted manner.

[0021] Step S3: Obtain the structural configuration data of the battery housing pressure relief valve; perform multi-valve linkage control topology analysis on the multi-dimensional time-series state characteristic data of the battery system using the structural configuration data of the battery housing pressure relief valve and the thermal runaway propagation risk characteristic data of the battery system, and generate multi-valve linkage control topology data of the battery system. In this embodiment of the invention, by pre-stored data on the structural configuration of the battery housing pressure relief valves in the database, the structural configuration information of all pressure relief valves on the battery housing is comprehensively collected, including the installation location, model specifications, burst pressure, response time, flow area, installation angle, and quantity distribution of the pressure relief valves. The coverage area, installation method, and core performance parameters of each pressure relief valve are clearly defined. All information is integrated to form battery housing pressure relief valve structural configuration data, providing basic parameter support for subsequent topology analysis. Based on this structural configuration data and the previously generated battery system thermal runaway propagation risk characteristic data, a multi-valve linkage control topology analysis is performed on the multi-dimensional time-series state characteristic data of the battery system. First, a pressure relief valve execution node model is established, with each pressure relief valve corresponding to an independent execution node, binding its structural configuration parameters and coverage area. The function association characteristics and collaborative relationships of each execution node are analyzed. Subsequently, the function characteristics of the pressure relief valve execution nodes are mapped and matched with the thermal runaway propagation risk characteristics to clarify the correspondence and function priority between different thermal runaway risks and each pressure relief valve. Candidate combinations of multi-valve linkage suitable for different risk scenarios are selected, and the optimal linkage combination is selected through constraint optimization. Finally, taking the execution node of the optimal linkage combination as the core, and combining the core parameters in the multi-dimensional time-series state characteristic data of the battery system, a multi-valve linkage control topology is constructed. The relationship, connection logic and triggering conditions of the control node and state node are clarified, and the multi-valve linkage control topology data of the battery system is generated. The control logic and node relationship of the multi-valve linkage are clearly presented, providing structured support for the subsequent design of the burst pressure release control strategy.

[0022] Step S4: Design the burst pressure release control strategy for the multi-valve linkage of the battery system based on the multi-valve linkage control topology data of the battery system, and generate the intelligent control strategy for burst pressure release of the multi-valve linkage of the battery system.

[0023] In this embodiment of the invention, a burst pressure release control strategy for a battery system with multi-valve linkage control is designed based on the multi-valve linkage control topology data. First, the burst pressure influencing factors of the multi-valve linkage control topology are analyzed, decomposing three core influencing factors: topology structure, valve performance, and battery system state parameters. The influence degree and action law of each factor on the burst pressure release effect are analyzed, generating multi-valve linkage burst pressure influencing factor data. Then, the heterogeneous combination characteristics of the multi-valve linkage are analyzed, dividing the heterogeneous combination dimensions according to valve type, installation location, and structural parameters. The core characteristics and synergistic effects of each heterogeneous combination are extracted, clarifying the suitable scenarios for different heterogeneous combinations, generating multi-valve linkage heterogeneous combination characteristic data. Based on the above two types of data, nonlinear time-series trend analysis of burst pressure influencing factors is conducted, tracking the change law and interrelationship of each core influencing factor over time, generating multi-valve linkage burst pressure influencing factor trend data. On this basis, dynamic prediction of burst pressure evolution is carried out based on this trend data, predicting the evolution trend, rate of change, and risk level changes of core burst pressure parameters such as battery system internal pressure and temperature, generating dynamic prediction data of burst pressure evolution. Finally, combining multi-valve linkage control topology data and dynamic prediction data of explosion pressure evolution, a hierarchical explosion pressure release control strategy is designed. The number of valves to be opened, the opening degree, the opening sequence and the trigger threshold are defined under different prediction periods and different risk scenarios. The closed-loop adjustment logic of the strategy is designed to ensure that the control strategy can dynamically adapt to the changes in explosion pressure evolution. All control parameters and logic are integrated to generate a multi-valve linkage intelligent control strategy for explosion pressure release in the battery system, so as to achieve accurate and efficient release of thermal runaway explosion pressure in the battery system and ensure the safety of the battery system and its surroundings.

[0024] Furthermore, step S1 includes the following steps: Step S11: Use a multi-type sensor array deployed inside the battery module and casing to monitor and process the battery system's operating status, and generate battery system operating status monitoring data; In this embodiment of the invention, multiple types of sensor arrays are arranged on the surface of each cell of the battery module, in the gap between modules, and in different areas inside the battery casing. The sensor array includes temperature sensors, gas sensors, internal pressure sensors, voltage and current sensors, and strain sensors. The temperature sensors are arranged at a density of 8 per module, located at the top, middle, and bottom of the cell. The gas sensors are arranged at the four corners inside the casing and at the center of the module. The internal pressure sensors are installed on the side wall and top of the casing through preset mounting holes. The voltage and current sensors are connected to the positive and negative terminals of the cell and the main circuit. The strain sensors are attached to the stress-prone parts of the casing. The sensor array collects various parameters during the operation of the battery system at a frequency of 100ms / time. The temperature sensor has a collection range of -40℃ to 150℃, the gas sensor focuses on collecting the real-time content of three gases: carbon monoxide, carbon dioxide, and hydrogen fluoride, the internal pressure sensor has a collection range of 0 to 1.2MPa, the voltage and current sensors have a collection range of 0 to 4.5V for voltage and -500A to 500A for current, and the strain sensor has a collection range of 0 to 5000με for casing strain. All collected parameters are integrated to form battery system operating status monitoring data, ensuring that the monitoring data can comprehensively cover the operating status of all key parts of the battery system.

[0025] Step S12: Perform multi-source response analysis of battery system operating status based on battery system operating status monitoring data to generate multi-source response data of battery system operating status; In this embodiment of the invention, based on the collected battery system operating status monitoring data, multi-source response analysis of the battery system operating status is carried out. Dynamic response extraction is performed on the temperature, gas concentration, internal pressure, voltage, current, and casing strain parameters in the monitoring data. The temperature dynamic response analysis uses a sliding window method, with a fixed window size of 5 acquisition cycles (500ms). Temperature monitoring data is segmented by segment through the sliding window, and the temperature change rate and peak temperature within each window are extracted. The temperature change rate is calculated as the ratio of the temperature difference between the beginning and end of the window to the window duration. The temperature peak is directly extracted as the highest temperature value within the window. The analysis results of all windows are integrated to form temperature dynamic response data, accurately reflecting the dynamic change law of the battery system temperature. The gas concentration change response analysis uses a difference calculation method, calculating the difference in concentrations of carbon monoxide, carbon dioxide, and hydrogen fluoride for each adjacent acquisition cycle. Simultaneously, the cumulative change in concentration of each gas from the start of monitoring to the current acquisition cycle is calculated. The concentration differences and cumulative changes of the three gases are recorded and integrated to form gas concentration change data. The thermal runaway response data clearly presents the release pattern of characteristic gases during thermal runaway. The internal pressure transient fluctuation response analysis uses a fluctuation amplitude calculation method to extract the maximum and minimum internal pressure values ​​within each acquisition cycle, calculating the difference as the internal pressure fluctuation amplitude for that cycle. Simultaneously, the time point of fluctuation occurrence is recorded. Fluctuation parameters from all cycles are integrated to form internal pressure transient fluctuation response data, capturing the instantaneous fluctuation characteristics of internal pressure. The voltage and current disturbance response analysis focuses on extracting the instantaneous abrupt changes in voltage and current values ​​and the duration of the disturbance. Clear judgment criteria are set: when a voltage abrupt change exceeds 0.1V, a current abrupt change exceeds 50A, and the duration exceeds 3 acquisition cycles (300ms), it is marked as a valid disturbance. The abrupt change amplitude, duration, and occurrence time of valid disturbances are recorded in detail. Disturbances that do not meet the judgment criteria are not included in valid data. All valid disturbance data are integrated to form voltage and current disturbance response data. The shell strain response analysis directly extracts the real-time strain change within each acquisition cycle, while simultaneously calculating the cumulative strain value from the start of monitoring to the current cycle. Both are recorded and integrated synchronously to form shell strain response data.

[0026] Step S13: Perform multi-physics coupling deconstruction processing on the battery system based on the multi-source response data of the battery system operating status to generate multi-physics coupling deconstruction data of the battery system; In this embodiment of the invention, based on the multi-source response data of the battery system's operating state, multi-physics coupling deconstruction processing of the battery system is carried out. It is clarified that the multi-physics field includes temperature, pressure, electric, and strain fields. A field coupling decomposition algorithm is used to strictly map each type of parameter in the multi-source response data to a different physical field, achieving precise division of the physical fields. Specifically, the temperature dynamic response data directly corresponds to the temperature field, comprehensively reflecting the spatial distribution and dynamic changes of the temperature field; the internal pressure transient fluctuation response data corresponds to the pressure field, capturing the transient fluctuations and spatial distribution characteristics of the pressure field; the voltage and current disturbance response data corresponds to the electric field, reflecting the disturbance changes and distribution patterns of the electric field; the shell strain response data corresponds to the strain field, presenting the spatial distribution and changing trend of the shell strain; and the gas concentration change response data serves as a coupling correlation parameter, connecting the four types of physical fields in series to reflect the mass transfer characteristics during the coupling process of each physical field. Based on the above parameter correspondences, a coupling correlation model between various physical fields is established. The model explicitly sets the coupling coefficients between each physical field: the coupling coefficient between the temperature field and the pressure field is set to 0.82, highlighting the impact of temperature rise on internal pressure; the coupling coefficient between the temperature field and the electric field is set to 0.65, reflecting the correlation between temperature changes and voltage / current disturbances; and the coupling coefficient between the pressure field and the strain field is set to 0.78, reflecting the influence of internal pressure fluctuations on shell strain. Through this coupling correlation model, the coupling components of each physical field are accurately decomposed, interference data between different physical fields are eliminated, and the independent characteristic parameters of each physical field are separated. Simultaneously, the coupling correlation parameters between the physical fields are retained. All independent characteristic parameters and coupling correlation parameters are categorized and integrated according to physical field type to form multi-physics coupling deconstruction data for the battery system, achieving accurate deconstruction of the multi-physics coupling state of the battery system.

[0027] Step S14: Perform battery system response decoupling characteristic analysis based on the multi-physics coupling deconstruction data of the battery system to generate battery system response decoupling characteristic data; In this embodiment of the invention, based on the multi-physics coupling deconstruction data of the battery system, the decoupling characteristic analysis of the battery system response is carried out. First, the response gradient distribution analysis of the multi-physics decoupling is performed on the coupling deconstruction data. The gradient descent algorithm is used, with a gradient step size of 0.01 and 100 iterations, to calculate the gradient distribution of the characteristic parameters of each physical field in space, thereby obtaining the multi-physics decoupling response gradient distribution data and clarifying the spatial variation law of the response of each physical field. Then, based on the gradient distribution data, the response peak value and dominant influencing parameters of each physical field in different regions are extracted. For the temperature field, the temperature change rate is the dominant parameter; for the pressure field, the internal pressure fluctuation amplitude is the dominant parameter; and for the electric field, the voltage disturbance amplitude is the dominant parameter. The variable field uses the cumulative strain as the dominant parameter to form the dominant characteristic data of the decoupled local response of the multi-physics field. At the same time, based on the decoupled data of the multi-physics field, the correlation coefficients between each physical field are calculated. The correlation coefficient between the temperature field and the pressure field is calculated using Pearson correlation analysis, and the correlation coefficients between the temperature field and the electric field, and between the pressure field and the strain field are calculated using Spearman correlation analysis, to obtain the decoupled correlation data of the multi-physics field and clarify the correlation strength between each physical field. Combining the dominant characteristic data of the local response and the correlation data, a feature fusion algorithm is used to fuse the independent features and correlation features of each physical field to obtain characteristic parameters that can reflect the independent effects and interrelationships of each physical field, thus forming the decoupled characteristic data of the battery system response.

[0028] Step S15: Perform spatial energy and matter migration mapping feature analysis using battery system response decoupling feature data to generate battery system energy-matter migration mapping feature data; In this embodiment of the invention, the spatial energy and matter migration mapping characteristics are analyzed by decoupling the characteristic data of the battery system response. The energy migration analysis focuses on the energy transfer between the temperature field and the electric field. Based on the characteristic parameters of the decoupled temperature field and electric field, the energy transfer rate per unit time is calculated for each acquisition cycle. The energy transfer rate calculation formula is strictly set as the ratio of the energy change to the time difference. The energy change is calculated as the sum of the temperature change and the electric field energy change. The time difference is the time interval (100ms) between two adjacent acquisition cycles. The energy transfer rate of different regions of the battery system is calculated point by point. The direction and intensity of energy transfer are recorded to clarify the migration law of energy from high temperature region to low temperature region and from high electric field intensity region to low electric field intensity region. Mass migration analysis focuses on the correlation between gas concentration changes and pressure fields. Based on decoupled gas concentration change data and pressure field characteristic parameters, a gas migration trajectory model is established. The model explicitly states a positive correlation between gas migration speed and internal pressure fluctuation amplitude, with a correlation coefficient set at 0.75. This means that the larger the internal pressure fluctuation amplitude, the faster the gas migration speed. Using this model, combined with gas concentration change data, the migration paths and distribution of carbon monoxide, carbon dioxide, and hydrogen fluoride within the battery casing are tracked periodically. The migration patterns from high-concentration areas to low-concentration areas and from high-pressure areas to low-pressure areas are clarified, while the concentration decay during gas migration is recorded. The energy transfer rate, direction, and intensity are integrated synchronously with the gas migration trajectory, distribution, and concentration decay data to form energy-mass migration mapping characteristic data of the battery system, clearly and comprehensively presenting the energy and mass migration patterns within the battery system.

[0029] Step S16: Based on the battery system response decoupling feature data and the battery system energy-mass migration mapping feature data, perform adaptive dimensionality reduction and resampling processing of the battery system response migration sensitive region to generate dimensionality reduction response decoupling feature data of the battery system. In this embodiment of the invention, based on the battery system response decoupling characteristic data and energy-mass migration mapping characteristic data, adaptive dimensionality reduction and resampling processing of the battery system's response migration sensitive region is carried out. Feature dimensionality reduction processing is performed, and a feature contribution rate analysis method is adopted to calculate the contribution rate of all feature parameters in the response decoupling characteristic data and energy-mass migration mapping characteristic data one by one. The feature contribution rate threshold is set to 0.05, and redundant feature parameters with contribution rates lower than this threshold are strictly eliminated. Core feature parameters with contribution rates higher than or equal to this threshold are retained. Core feature parameters include temperature change rate, internal pressure fluctuation amplitude, voltage disturbance amplitude, cumulative strain value, energy transfer rate, gas migration speed, and three characteristic gas concentrations. Through feature dimensionality reduction, the amount of data is greatly reduced, while retaining the core features that can reflect the state of the battery system, simplifying the subsequent analysis process. Response migration sensitive areas were located and adaptively resampled. Sensitive area criteria were clearly defined, strictly defined as areas with energy transfer rates exceeding 50 J / s, gas migration velocities exceeding 0.2 m / s, and cumulative strain exceeding 1000 με. By comparing the characteristic parameters of all areas with the criteria, the response migration sensitive areas were accurately located. These sensitive areas are mainly concentrated in the module center, around the pressure relief valve, and in stress-prone parts of the casing. These areas are critical for thermal runaway propagation and are also key areas for multi-valve linkage pressure release control. For the characteristic data of sensitive areas, a higher-frequency resampling method was used, increasing the resampling frequency from the conventional 100 ms / time to 50 ms / time, ensuring accurate capture of instantaneous changes in the characteristic parameters of sensitive areas without missing any critical data. For the characteristic data of non-sensitive areas, a conventional resampling method was used, maintaining a resampling frequency of 100 ms / time, further reducing the data volume while ensuring data accuracy. Through this adaptive adjustment of the resampling frequency, the higher-frequency resampling data of sensitive areas and the conventional resampling data of non-sensitive areas were integrated to ultimately form the decoupled characteristic data of the battery system's response.

[0030] Step S17: Perform multi-dimensional time-series state feature analysis on the battery system distribution dimensionality reduction of the battery system response decoupling feature data and the battery system energy-mass migration mapping feature data through the battery system dimensionality reduction response decoupling feature data, and generate multi-dimensional time-series state feature data of the battery system.

[0031] In this embodiment of the invention, multi-dimensional time-series state feature analysis of the battery system distribution is performed by using the battery system's dimensionality-reduced response decoupling feature data and energy-mass migration mapping feature data. A time-series feature extraction algorithm is employed, with a time-series window size of 10 acquisition cycles. The dimensionality-reduced response decoupling feature data is segmented within the time-series window, and the changing trend, peaks, valleys, and changing cycles are extracted window by window. The changing trend is calculated based on the slope of the data within the window. Peaks and valleys are directly extracted as the maximum and minimum values ​​within the window. The changing cycle is calculated based on the time interval between two adjacent peaks or valleys, comprehensively capturing the time-series change patterns of the dimensionality-reduced response decoupling feature data. Simultaneously, the extracted time-series features are correlated and matched with the energy transfer and mass migration parameters in the energy-mass migration mapping feature data to establish a multi-dimensional time-series state feature model. The model explicitly sets the weight of the time-series features at 0.6 and the weight of the energy-mass migration features at 0.4. The weight allocation is based on the importance of the two types of features in determining the battery system state. The time-series features primarily reflect the dynamic changes of the battery system, while the energy-mass migration features primarily reflect the potential risk of thermal runaway. This model integrates time-series features with energy-matter migration features in a weighted manner to obtain parameters that comprehensively reflect the real-time operating status, time-series change patterns, and energy-matter migration characteristics of the battery system. All parameters are categorized and integrated according to the acquisition timestamp and spatial region to form multi-dimensional time-series state feature data of the battery system. This provides accurate and efficient data support for subsequent thermal runaway propagation risk analysis and multi-valve linkage control, ensuring that multi-valve linkage explosion pressure release control can accurately adapt to the dynamic operating status of the battery system.

[0032] Furthermore, the multi-source response data of the battery system operating status mentioned in step S12 includes temperature dynamic response data, gas concentration change response data, internal pressure transient fluctuation response data, voltage and current disturbance response data, and casing strain response data.

[0033] Furthermore, step S14 includes the following steps: Step S141: Perform multi-physics decoupling response gradient distribution analysis on the multi-physics coupling deconstruction data of the battery system to generate multi-physics decoupling response gradient distribution data; In this embodiment of the invention, response gradient distribution analysis of multi-physics coupling deconstruction data of a battery system is carried out. The multi-physics coupling deconstruction data includes independent characteristic parameters of temperature field, pressure field, electric field, strain field and coupling correlation parameters between each field. The analysis adopts gradient descent algorithm, with a gradient step size of 0.01, 100 iterations, and convergence error controlled within 0.001. During the analysis, using the three-dimensional coordinates inside the battery system casing as a reference, the coupled deconstruction data was divided into several 10mm×10mm×10mm analysis units according to spatial regions. The gradient values ​​of the characteristic parameters of each physical field within each analysis unit were calculated. The temperature field gradient calculation used the rate of temperature change as the core parameter, the pressure field gradient calculation used the amplitude of internal pressure fluctuation as the core parameter, the electric field gradient calculation used the amplitude of voltage disturbance as the core parameter, and the strain field gradient calculation used the cumulative strain as the core parameter. At the same time, the spatial distribution and trend of each gradient value were recorded. The gradient calculation results and spatial distribution information of all analysis units were integrated to form multi-physics decoupling response gradient distribution data, which clearly presents the spatial gradient change law of the response characteristics of each physical field after decoupling within the battery system.

[0034] Step S142: Perform dominant feature analysis of local response of multiphysics decoupling based on the gradient distribution data of multiphysics decoupling response, and generate dominant feature data of local response of multiphysics decoupling response; In this embodiment of the invention, the dominant feature analysis of the local response of multi-physics decoupling is carried out based on the gradient distribution data of the multi-physics decoupling response. First, the gradient distribution data is divided into regions, and the battery system is divided into three core local regions: the cell region, the module gap region, and the casing region. Each region is further subdivided into several sub-regions. For each sub-region, the gradient peak value and corresponding feature parameters of each physical field in that region are extracted. Gradient peak value thresholds are set: 5℃ / mm for temperature field gradient peak value, 0.05MPa / mm for pressure field gradient peak value, 0.1V / mm for electric field gradient peak value, and 50με / mm for strain field gradient peak value. Sub-regions with gradient peak values ​​exceeding the corresponding thresholds are selected as key regions of local response. For key areas, the dominant characteristic parameters of the local response of each physical field are determined. For the temperature field, the rate of temperature change is used as the dominant characteristic parameter; for the pressure field, the amplitude of internal pressure fluctuation is used as the dominant characteristic parameter; for the electric field, the amplitude of voltage disturbance is used as the dominant characteristic parameter; and for the strain field, the cumulative strain is used as the dominant characteristic parameter. At the same time, the numerical range, rate of change, and distribution of each dominant characteristic parameter in the local area are recorded. For non-key areas, only the average value of the core dominant characteristic parameter is recorded. The dominant characteristic information of all local areas is integrated to form multi-physics field decoupled local response dominant characteristic data, and the core response characteristics of each local area are clarified.

[0035] Step S143: Perform correlation analysis on the multi-physics decoupling based on the multi-physics coupling deconstruction data of the battery system, and generate multi-physics decoupling correlation data; In this embodiment of the invention, a correlation analysis of multi-physics field decoupling is conducted based on the multi-physics field coupling deconstruction data of the battery system. The analysis objects are the independent characteristic parameters of the temperature field, pressure field, electric field, and strain field. A scenario-based correlation analysis method is adopted, dividing the data into three operating scenarios: normal operation of the battery system, slight disturbance, and abnormal temperature rise. The correlation coefficients between different physical fields in each scenario are calculated separately. The correlation coefficient between the temperature field and the pressure field is calculated using Pearson correlation analysis, with the temperature change rate and internal pressure fluctuation amplitude as the core correlation parameters. The calculation window is set to 20 acquisition cycles, and the correlation coefficient is calculated once in each window. The average of the correlation coefficients of all windows is taken as the final correlation coefficient. The correlation coefficients between the temperature field and the electric field, and between the pressure field and the strain field are calculated using Spearman correlation analysis. The temperature field and the electric field use the temperature change rate and voltage disturbance amplitude as the core parameters, and the pressure field and the strain field use the internal pressure fluctuation amplitude and cumulative strain as the core parameters. The calculation window is also set to 20 acquisition cycles, and the correlation coefficient of each window is calculated and averaged. By integrating the correlation coefficients, correlation parameters, and correlation trends between different physical fields in various scenarios, multi-physics decoupled correlation data is formed, clarifying the correlation strength and correlation rules between each physical field, and avoiding interference between different physical fields.

[0036] Step S144: Perform battery system response decoupling feature analysis using multi-physics decoupling local response dominant feature data and multi-physics decoupling correlation data to generate battery system response decoupling feature data.

[0037] In this embodiment of the invention, battery system response decoupling feature analysis is carried out using multi-physics decoupled local response dominant feature data and multi-physics decoupled correlation data. A feature fusion algorithm is adopted, setting the weight of the local response dominant feature to 0.7 and the weight of the multi-physics correlation feature to 0.3, to achieve accurate fusion of the two types of data. During the fusion process, the dominant parameters of each physics field in the local response dominant feature data are first classified by region. Then, the correlation coefficients in the multi-physics decoupled correlation data are combined to perform correlation calibration on the dominant parameters of different physics fields. When the correlation coefficient between two physics fields is greater than 0.7, their corresponding dominant parameters are fused collaboratively, and the calibration coefficient is set to 0.85. When the correlation coefficient is between 0.3 and 0.7, an independent fusion method is used to retain the core features of each dominant parameter. When the correlation coefficient is less than 0.3, weakly correlated parameters are removed, and only the core dominant parameters are retained. After fusion, the independent decoupling features and inter-field correlation decoupling features of each physical field are extracted. For each physical field, 3-5 core decoupling feature parameters are selected. The temperature field retains the temperature change rate, gradient peak, and regional mean; the pressure field retains the internal pressure fluctuation amplitude, gradient peak, and fluctuation frequency; the electric field retains the voltage disturbance amplitude and disturbance duration; and the strain field retains the cumulative strain and strain change rate. All core decoupling feature parameters and correlation feature information are integrated to form battery system response decoupling feature data. This data can accurately reflect the independent effects and interrelationships of each physical field.

[0038] Furthermore, step S2 includes the following steps: Step S21: Perform multi-layer heterogeneous characteristic analysis of the battery system based on the multi-dimensional time-series state characteristic data of the battery system, generate multi-layer heterogeneous characteristic data of the battery system, and perform multi-layer heterogeneous structural node modeling processing through the multi-layer heterogeneous characteristic data of the battery system to generate multi-layer heterogeneous structural node data. In this embodiment of the invention, multi-layer heterogeneous characteristic analysis of the battery system is carried out based on multi-dimensional time-series state characteristic data of the battery system. The multi-dimensional time-series state characteristic data includes the decoupling characteristics of temperature field, pressure field, electric field, and strain field, as well as energy-matter migration mapping characteristics. The analysis adopts a hierarchical analytical algorithm to divide the battery system into three heterogeneous layers: cell layer, module layer, and shell layer. The cell layer uses a single cell as the basic unit, the module layer uses each group of battery modules as the basic unit, and the shell layer uses the entire battery shell as the basic unit. For each layer, the structural characteristics, time-series variation characteristics, and energy-matter migration characteristics of that layer are extracted. The cell layer focuses on extracting the temperature change rate and voltage disturbance amplitude of a single cell, the module layer focuses on extracting the energy transfer rate and gas migration velocity between cells within a module, and the shell layer focuses on extracting the cumulative value of shell strain and the overall internal pressure fluctuation characteristics. The feature parameters of the three layers are integrated to form multi-layer heterogeneous characteristic data of the battery system. Based on this data, multi-layer heterogeneous structural node modeling was carried out. The core feature parameters of each level were used as nodes. The cell layer nodes were set with one core node corresponding to each cell, the module layer nodes were set with one summary node corresponding to each module, and the shell layer was set with one overall node. The nodes were connected according to the hierarchical relationship. The cell layer nodes were connected to the corresponding module layer nodes, and all module layer nodes were connected to the shell layer nodes. The node connection weight was set according to the hierarchical importance. The connection weight between the cell layer and the module layer nodes was 0.6, and the connection weight between the module layer and the shell layer nodes was 0.4. After the modeling was completed, multi-layer heterogeneous structural node data was generated, which clearly presented the structural relationship and feature distribution of each level of the battery system.

[0039] Step S22: Analyze the multi-mechanism propagation relationship of the battery system based on the multi-layer heterogeneous structure node data, generate multi-mechanism propagation relationship data of the battery system, and analyze the multi-mechanism propagation driving factors through the multi-mechanism propagation relationship data of the battery system to generate multi-mechanism propagation driving factor data. In this embodiment of the invention, multi-mechanism propagation relationship analysis of the battery system is carried out based on multi-layer heterogeneous structure node data. Multi-mechanism propagation includes three core propagation mechanisms: thermal conduction, gas diffusion, and electrical conduction. The analysis employs a mechanism decomposition and correlation method to analyze the propagation relationship between nodes under each mechanism. For the thermal conduction mechanism, the temperature change rate is used as the core propagation parameter. The temperature difference between adjacent nodes is calculated. When the temperature difference is greater than 3°C, a thermal conduction propagation relationship is determined between the two nodes, and the propagation intensity is calculated as the ratio of the temperature difference to the distance. For the gas diffusion mechanism, the gas migration velocity is used as the core propagation parameter. When the gas migration velocity between nodes is greater than 0.1 m / s, a gas diffusion propagation relationship is determined, and the propagation intensity is set according to the gas concentration difference. For the electrical conduction mechanism, the voltage perturbation amplitude is used as the core propagation parameter. When the voltage perturbation amplitude difference between adjacent nodes is greater than 0.05V, an electrical conduction propagation relationship is determined, and the propagation intensity is set according to the ratio of the voltage perturbation amplitudes. The node propagation relationships, propagation intensities, and propagation directions under the three propagation mechanisms are integrated to form multi-mechanism propagation relationship data for the battery system. Based on this data, a multi-mechanism propagation driving factor analysis was conducted. The driving factors included four categories: temperature gradient, pressure difference, voltage disturbance intensity, and gas concentration gradient. The factor contribution rate calculation method was adopted, with a calculation window of 15 acquisition cycles. The contribution rate of each of the four factors to each propagation mechanism was calculated. The contribution rate of temperature gradient to the heat conduction mechanism was calculated using the difference ratio method, the contribution rate of pressure difference to the gas diffusion mechanism was calculated using the fluctuation correlation method, the contribution rate of voltage disturbance intensity to the electrical conduction mechanism was calculated using the amplitude ratio method, and the contribution rate of gas concentration gradient to the gas diffusion mechanism was calculated using the gradient ratio method. Factors with a contribution rate greater than 0.2 were extracted as core driving factors. The numerical range, trend of change, and degree of influence on the propagation mechanism of each core driving factor were recorded to form multi-mechanism propagation driving factor data, clarifying the core driving force of thermal runaway propagation and providing accurate basis for subsequent propagation path analysis.

[0040] Step S23: Analyze the propagation path distribution characteristics based on the multi-layer heterogeneous structure node data and the multi-mechanism propagation driving factor data to generate propagation path distribution characteristic data; In this embodiment of the invention, propagation path distribution characteristics are analyzed based on multi-layer heterogeneous structure node data and multi-mechanism propagation driving factor data. A path tracing algorithm is employed, using multi-layer heterogeneous structure nodes as tracking nodes and multi-mechanism propagation driving factors as the tracking basis. The tracking step size is set to one acquisition cycle, starting from the cell layer node, to track the propagation trajectory between nodes under the influence of each core driving factor. Thermal conduction propagation path tracking uses temperature gradient as the core basis, prioritizing the tracking of node connection paths with a temperature gradient greater than 2℃ / mm; gas diffusion propagation path tracking uses pressure difference as the core basis, prioritizing the tracking of node connection paths with a pressure difference greater than 0.03MPa; and electrical conduction propagation path tracking uses voltage disturbance intensity as the core basis, prioritizing the tracking of node connection paths with a voltage disturbance intensity greater than 0.08V. For each tracked propagation path, record all nodes on the path, propagation mechanism type, propagation speed, and core driving factor values. At the same time, calculate the propagation length and propagation time of each path. Paths with a propagation length greater than 50 mm and a propagation time less than 10 s are marked as key propagation paths. Integrate all relevant information of propagation paths, including path distribution location, propagation mechanism, propagation parameters, and key markings, to form propagation path distribution characteristic data, presenting the propagation path and distribution pattern of thermal runaway at each level and node of the battery system.

[0041] Step S24: Analyze the risk clustering characteristics of the transmission path based on the distribution characteristic data of the transmission path, and generate risk clustering characteristic data of the transmission path; In this embodiment of the invention, risk clustering characteristic analysis of propagation paths is conducted based on propagation path distribution characteristic data. A risk clustering degree calculation method is used, with each node on each propagation path as the analysis unit. The risk clustering degree of each node is calculated using the formula: the sum of the core driving factor values ​​multiplied by the propagation intensity. A risk clustering degree threshold of 80 is set, and nodes with a risk clustering degree exceeding this threshold are marked as risk clustering nodes. For each propagation path, the number, distribution location, and risk clustering degree values ​​of risk clustering nodes on the path are statistically analyzed. The average risk clustering degree of each path is calculated using the arithmetic mean method, and paths with an average risk clustering degree greater than 60 are marked as high-risk clustering paths. Simultaneously, the correlation between risk cluster nodes is analyzed. When the propagation distance between two risk cluster nodes is less than 30mm and the propagation time is less than 5s, a risk cluster area is determined to be formed. The range of the risk cluster area, the nodes it contains, and the core driving factors are recorded. All relevant information of risk cluster nodes, high-risk cluster paths, and risk cluster areas is integrated to form risk cluster characteristic data of propagation paths. The location, degree, and pattern of risk clustering during the thermal runaway propagation process are clarified, and high-risk areas of thermal runaway are accurately located.

[0042] Step S25: Analyze the propagation risk characteristics of the battery system through the propagation path risk clustering feature data, and generate the thermal runaway propagation risk feature data of the battery system.

[0043] In this embodiment of the invention, the thermal runaway propagation risk characteristics of the battery system are analyzed using propagation path risk aggregation characteristic data. A risk level classification method is adopted, combining the propagation speed and length from the propagation path distribution characteristic data with the risk aggregation degree from the risk aggregation characteristic data, to classify the thermal runaway propagation risk into three levels: low risk level corresponds to an average risk aggregation degree of less than 40 and a propagation speed of less than 0.05 m / s; medium risk level corresponds to an average risk aggregation degree between 40 and 60 and a propagation speed between 0.05 m / s and 0.1 m / s; and high risk level corresponds to an average risk aggregation degree greater than 60 and a propagation speed greater than 0.1 m / s. For each risk level, the corresponding propagation path, risk aggregation area, core driving factor, and risk evolution trend are recorded. Simultaneously, the coverage area of ​​each risk level is calculated, based on the proportion of the risk area to the total volume of the battery system. The coverage area for the low risk level is less than 20%, the coverage area for the medium risk level is between 20% and 50%, and the coverage area for the high risk level is greater than 50%. Furthermore, the core characteristics of thermal runaway propagation under different risk levels were analyzed. Low-risk levels are dominated by a single propagation mechanism, medium-risk levels are dominated by two propagation mechanisms working together, and high-risk levels are dominated by three propagation mechanisms working together. By integrating the risk level classification results, the characteristics of each level, the coverage, and the evolution trend, risk characteristic data of thermal runaway propagation in battery systems were formed, which comprehensively and accurately reflects the risk status of thermal runaway propagation in battery systems.

[0044] Furthermore, as an embodiment of the present invention, reference is made to... Figure 2 As shown, Figure 1 A detailed flowchart illustrating the implementation steps of step S3 is provided in this embodiment. Step S3 includes: Step S31: Obtain the structural configuration data of the battery housing pressure relief valve; In this embodiment of the invention, the structural configuration data of the battery housing pressure relief valve is acquired. Six pressure relief valves are arranged on the battery housing, with two installed on the top and four on the side wall. The top pressure relief valve is model XY-01, and the side wall pressure relief valve is model XY-02. All pressure relief valves adopt a burst-type structure, with a burst pressure set to 0.8MPa, an opening response time ≤50ms, a flow area of ​​150mm², and an installation angle of 30° with the surface of the housing. The acquired data covers the installation location, model specifications, burst pressure, response time, flow area, installation angle, and quantity distribution of pressure relief valves. It also records the corresponding housing area for each pressure relief valve: the two top pressure relief valves correspond to the middle and rear areas of the battery system, respectively, while the four side pressure relief valves correspond to the four side areas of the battery module. This clarifies the coverage area and installation method of each pressure relief valve. Integrating all the above information forms the battery housing pressure relief valve structural configuration data, providing basic parameters for subsequent mapping analysis of thermal runaway risk and valve function. This ensures that multi-valve linkage control can match the actual structure of the pressure relief valves, guaranteeing the accuracy and effectiveness of burst pressure release, and meeting the core requirements of multi-valve linkage burst pressure release.

[0045] Step S32: Based on the battery casing pressure relief valve structure configuration data and battery system thermal runaway propagation risk characteristic data, perform thermal runaway risk and valve action mapping relationship analysis to generate thermal runaway risk-valve action mapping relationship data; In this embodiment of the invention, based on the structural configuration data of the battery casing pressure relief valve and the characteristic data of thermal runaway propagation risk of the battery system, an analysis of the mapping relationship between thermal runaway risk and valve function is conducted. A mapping matching algorithm is used. First, pressure relief valve execution nodes are modeled based on the structural configuration data. Each pressure relief valve corresponds to one execution node. Node parameters include the installation position of the pressure relief valve, burst pressure, flow area, and coverage range. During modeling, each execution node is bound to the corresponding battery system area to clarify the scope of action of each node. The function association characteristics of the pressure relief valve execution nodes are analyzed, and the pressure relief coverage radius of each execution node is calculated. The coverage radius is calculated as the product of the square root of the flow area and the installation height. The coverage radius of the top pressure relief valve is 150mm, and the coverage radius of the side wall pressure relief valve is 120mm. Simultaneously, the pressure relief efficiency of each execution node is determined. The pressure relief efficiency is directly proportional to the flow area and inversely proportional to the response time, calculated as the flow area divided by the response time. The relationship between thermal runaway propagation risk characteristics and data of battery system is abstracted, extracting risk levels, risk clustering areas, and core driving factors. High, medium, and low risk levels are assigned different pressure thresholds: high risk corresponds to internal pressure ≥ 0.6 MPa, medium risk to internal pressure 0.4-0.6 MPa, and low risk to internal pressure < 0.4 MPa. The correlation data of pressure relief valve actuators are mapped to the abstract characteristic data of thermal runaway propagation risk. Based on the coverage and pressure relief efficiency of each actuator, the corresponding risk area and risk level are matched. When the overlap area between the actuator's coverage and the risk clustering area is ≥ 60%, a mapping relationship is determined between the actuator and the corresponding risk. The risk level, risk clustering area, and core driving factors corresponding to each actuator are recorded, and integrated to form thermal runaway risk-valve action mapping data. This clarifies the correspondence between different thermal runaway risks and the actions of each pressure relief valve, providing a basis for subsequent multi-valve linkage combination analysis.

[0046] Step S33: Based on the thermal runaway risk-valve action mapping relationship data, perform multi-valve linkage candidate combination analysis for thermal runaway risk to generate multi-valve linkage candidate combination data; perform constraint optimization solution on the multi-valve linkage candidate combination data to generate multi-valve linkage combination optimization data; In this embodiment of the invention, based on the thermal runaway risk-valve action mapping data, a multi-valve linkage candidate combination analysis is conducted. A combination screening algorithm is used to divide combination scenarios according to the thermal runaway risk level: high-risk scenarios correspond to 3-4 pressure relief valve linkages, medium-risk scenarios to 2-3 pressure relief valve linkages, and low-risk scenarios to 1-2 pressure relief valve linkages. Within each scenario, pressure relief valve execution nodes corresponding to the risk area are selected based on the mapping data. These nodes are then combined according to their overlap in coverage area; execution nodes with an overlap of ≤30% are included in the same candidate combination to avoid efficiency waste caused by repeated pressure relief ranges. Simultaneously, it is ensured that the total flow area of ​​the pressure relief valves within the combination meets the pressure relief requirements of the corresponding risk level: high-risk scenarios have a total flow area ≥600mm², medium-risk scenarios ≥450mm², and low-risk scenarios ≥150mm². For each risk scenario, 5-8 candidate combinations are selected, and the number, model, installation location, total flow area, and covered risk area of ​​each combination are recorded to form multi-valve linkage candidate combination data. Constraint optimization of multi-valve linkage candidate combinations was performed using a constraint optimization algorithm. The constraints were set as follows: response time ≤ 50ms, total flow area meeting the corresponding risk level requirements, and pressure relief coverage ≥ 90% of the corresponding risk area. The optimization objective was to maximize pressure relief efficiency. The optimization step size was set to 0.02, and the number of iterations was 80. The pressure relief efficiency and constraint satisfaction of each candidate combination were calculated. Combinations that did not meet the constraints were eliminated, and the combination with the highest pressure relief efficiency was selected as the optimal combination. All parameters and optimization process data of the optimal combination were recorded to form multi-valve linkage combination optimization data. This ensures that multi-valve linkage combinations can efficiently adapt to corresponding thermal runaway risks and improve pressure release efficiency.

[0047] Step S34: Based on the multi-valve linkage combination optimization data, perform multi-valve linkage control topology analysis on the multi-dimensional time-series state characteristic data of the battery system to generate multi-valve linkage control topology data of the battery system.

[0048] In this embodiment of the invention, based on multi-valve linkage combination optimization data, a multi-valve linkage control topology analysis of the battery system's multi-dimensional time-series state characteristic data is performed. A topology modeling algorithm is employed, using the pressure relief valve execution node in the multi-valve linkage combination optimization data as the core control node, and core parameters such as temperature, internal pressure, and strain in the battery system's multi-dimensional time-series state characteristic data as state nodes, to construct a multi-valve linkage control topology. Control nodes and state nodes are connected according to their association relationships, with each control node corresponding to multiple state nodes. State node parameters include the temperature change rate, internal pressure fluctuation amplitude, and cumulative strain value. The connection weights between control nodes and state nodes are set according to the degree of influence of the state node parameters on the opening of the pressure relief valve: the weight for internal pressure fluctuation amplitude is 0.5, the weight for temperature change rate is 0.3, and the weight for cumulative strain value is 0.2. A control logic node is set in the topology to receive real-time parameters from the status node. When the status node parameter reaches the corresponding threshold, the pressure relief valve corresponding to the control node is triggered to open. The internal pressure threshold is set according to the risk level: 0.6 MPa for high risk, 0.4 MPa for medium risk, and 0.2 MPa for low risk. The temperature change rate threshold is 5℃ / s, and the cumulative strain threshold is 2000με. Simultaneously, linkage logic is set in the topology to clarify the opening sequence of each pressure relief valve in the optimal combination. In high-risk scenarios, the side wall pressure relief valve is opened first, followed by the top pressure relief valve after a 20ms interval. In medium- and low-risk scenarios, the side wall pressure relief valve in the corresponding risk area is opened first, followed by the top pressure relief valve based on pressure changes. Integrating the control node, status node, connection weights, and linkage logic forms the multi-valve linkage control topology data for the battery system. This clearly presents the control logic, node associations, and opening sequence of the multi-valve linkage, providing structured support for the subsequent design of the burst pressure release control strategy. This ensures that the multi-valve linkage burst pressure release can accurately respond to the real-time status of the battery system and achieve efficient pressure relief.

[0049] Furthermore, step S32 includes the following steps: Step S321: Based on the battery housing pressure relief valve structure configuration data, perform pressure relief valve execution node modeling to generate battery system pressure relief valve execution node data; In this embodiment of the invention, pressure relief valve execution node modeling is performed based on the battery housing pressure relief valve structural configuration data. The battery housing pressure relief valve structural configuration data includes the installation position, model specifications, burst pressure, response time, flow area, and coverage range of six pressure relief valves. Among them, the top two pressure relief valves are model XY-01, with a burst pressure of 0.8MPa, a response time of 50ms, a flow area of ​​150mm², and are installed in the middle and rear areas of the battery system, with a coverage radius of 150mm. The four side wall pressure relief valves are model XY-02, with a burst pressure of 0.8MPa, a response time of 45ms, a flow area of ​​120mm², and are installed in the four side areas of the battery module, with a coverage radius of 120mm. The modeling employs a node mapping method, treating each pressure relief valve as an independent execution node. Each node is assigned a unique identifier: top pressure relief valves are identified as F1 and F2, while side wall pressure relief valves are identified as F3, F4, F5, and F6. Each execution node is bound to its corresponding structural configuration parameters, which explicitly include the identifier, installation coordinates, model, burst pressure, response time, flow area, and coverage range. Simultaneously, each node is associated with a corresponding area of ​​the battery system: F1 corresponds to the central area, F2 to the tail area, and F3-F6 to the four side areas of the module, respectively. After modeling, the identifiers, parameters, and associated area information of all execution nodes are integrated to generate battery system pressure relief valve execution node data, clearly presenting the execution node attributes and corresponding areas of each pressure relief valve.

[0050] Step S322: Perform a function correlation feature analysis on the pressure relief valve execution node data of the battery system to generate function correlation feature data of the pressure relief valve execution node; In this embodiment of the invention, the function association feature analysis of the pressure relief valve execution node data of the battery system is performed. A feature extraction algorithm is used, with each pressure relief valve execution node as the analysis unit, to extract the function association parameters of the node, including pressure relief coverage, pressure relief efficiency, response sensitivity and regional correlation. The pressure relief coverage area is calculated based on the circular area defined by the coverage radius of the execution node. The center of the circular area is the installation center of the pressure relief valve, and the radius is determined by the product of the square root of the flow area and the installation height. The coverage radius of F1 and F2 is 150mm, and the coverage radius of F3-F6 is 120mm. The pressure relief efficiency is calculated by dividing the flow area by the response time. The pressure relief efficiency of F1 and F2 is 3mm² / ms, and the pressure relief efficiency of F3-F6 is 2.67mm² / ms. The response sensitivity is set according to the ratio of burst pressure to response time, and the response sensitivity of all execution nodes is 0.016MPa / ms. The regional correlation is calculated based on the percentage of overlap between the coverage area of ​​the execution node and the corresponding battery system area. The regional correlation of F1 and F2 is 80%, and the regional correlation of F3-F6 is 75%. Simultaneously, the interaction relationships between each execution node are analyzed. When the overlapping area of ​​the coverage areas of two execution nodes is ≥30%, a synergistic relationship is determined. The parameters of the synergistic node pairs and the overlapping area are recorded. The interaction relationship parameters, synergistic relationships and regional association information of all execution nodes are integrated to generate the interaction relationship feature data of the pressure relief valve execution nodes. The action capability and synergistic relationship of each pressure relief valve execution node are clarified, providing a basis for subsequent risk and valve action mapping.

[0051] Step S323: Perform relational abstraction processing on the thermal runaway propagation risk characteristic data of the battery system to generate abstract characteristic data of thermal runaway propagation risk; In this embodiment of the invention, the thermal runaway propagation risk characteristic data of the battery system is subjected to relational abstraction processing. The thermal runaway propagation risk characteristic data includes three risk levels: high, medium, and low. Each level corresponds to different risk clustering areas, propagation speeds, and core driving factors. The high-risk level has an average risk clustering degree of over 60, a propagation speed of over 0.1 m / s, and a coverage area of ​​over 50%. The core driving factors are temperature gradient and pressure difference. The medium-risk level has an average risk clustering degree of 40 to 60, a propagation speed of 0.05 m / s to 0.1 m / s, and a coverage area of ​​20% to 50%. The core driving factor is temperature gradient. The low-risk level has an average risk clustering degree of less than 40, a propagation speed of less than 0.05 m / s, and a coverage area of ​​less than 20%. The core driving factor is voltage disturbance intensity. The abstract processing employs a feature extraction method to eliminate redundant descriptions of risk evolution trends, retaining core characteristic parameters for each risk level, including risk level identifiers, average risk clustering thresholds, propagation speed thresholds, coverage thresholds, and core driving factors. Simultaneously, the risk clustering areas corresponding to each risk level are delineated using three-dimensional coordinates, clarifying the spatial range of each risk area and the battery system area it encompasses. High, medium, and low risk levels are labeled R1, R2, and R3, respectively. By integrating risk level identifiers, core parameters, and risk area boundary information, abstract characteristic data of thermal runaway propagation risk is generated, simplifying the risk characteristic description and highlighting core risk parameters.

[0052] Step S324: Map the function-related feature data of the pressure relief valve to the abstract feature data of thermal runaway propagation risk to perform thermal runaway risk and valve function mapping relationship analysis, and generate thermal runaway risk-valve function mapping relationship data.

[0053] In this embodiment of the invention, the function-related feature data of the pressure relief valve execution node is mapped to the abstract feature data of thermal runaway propagation risk. The mapping relationship between thermal runaway risk and valve function is analyzed. A mapping matching algorithm is used, with the risk level and risk area in the abstract feature data of thermal runaway propagation risk as the matching target and the execution node parameters in the function-related feature data of the pressure relief valve execution node as the matching basis. A matching threshold is set, and a successful match is determined when the overlap area between the risk area and the execution node coverage area is ≥60%. During the matching process, for R1-level risk areas, priority is given to matching execution nodes with a pressure relief efficiency ≥3mm² / ms and a coverage radius ≥150mm, namely F1 and F2. At the same time, execution nodes corresponding to the core driving factors of R1 are also matched. For R1 areas driven by temperature gradient, all execution nodes are matched. For R1 areas driven by pressure difference, F3-F6 are matched with priority. For R2-level risk areas, execution nodes with a pressure relief efficiency of 2.5-3mm² / ms and a coverage radius of 120-150mm are matched, namely all execution nodes. Priority is given to matching the side wall execution nodes F3-F6 of the corresponding risk area. For R3-level risk areas, execution nodes with a pressure relief efficiency ≤2.67mm² / ms and a coverage radius ≤120mm are matched, namely F3-F6. For each successfully matched combination, the risk level, risk area, execution node identifier, matching parameters, and action priority are recorded. The action priority is set according to the product of pressure relief efficiency and area correlation. When the product is ≥2.4, the priority is high; when it is 2.0-2.4, it is medium; and when it is ≤2.0, it is low. All matching combinations, action priorities, and correlation parameters are integrated to generate thermal runaway risk-valve action mapping relationship data, clarifying the correspondence and action priority between different thermal runaway risks and each pressure relief valve execution node.

[0054] Furthermore, step S4 includes the following steps: Step S41: Analyze the factors affecting the burst pressure of the multi-valve linkage control topology based on the multi-valve linkage control topology data of the battery system, and generate multi-valve linkage burst pressure influencing factor data; In this embodiment of the invention, an analysis of the factors influencing the burst pressure of the multi-valve linkage control topology is conducted based on the multi-valve linkage control topology data of the battery system. The multi-valve linkage control topology data includes control nodes, state nodes, connection weights, and linkage logic. The control nodes correspond to six pressure relief valve execution nodes (identified as F1-F6), and the state nodes correspond to parameters such as temperature change rate, internal pressure fluctuation amplitude, and cumulative strain value. The linkage logic specifies the opening sequence and trigger threshold of each pressure relief valve. The analysis employs a factor decomposition method, classifying the factors influencing burst pressure into three categories: topology factors, valve-specific factors, and state parameter factors. Topology factors include the connection weights between control nodes and state nodes, and the linkage logic interval time. Valve-specific factors include the burst pressure of the pressure relief valve, response time, and flow area. State parameter factors include the internal pressure fluctuation amplitude, temperature change rate, and cumulative strain value. Quantitative analysis was performed on each type of factor. For topology factors, connection weights were set at 0.5 for internal pressure, 0.3 for temperature, and 0.2 for strain, with a linkage logic interval of 20ms. For valve-specific factors, burst pressure was 0.8MPa, response time was 45-50ms, and flow area was 120-150mm². For state parameter factors, internal pressure thresholds were 0.2-0.6MPa, temperature change rate thresholds were 5℃ / s, and cumulative strain thresholds were 2000με. The influence coefficients of each factor on the burst pressure release effect were calculated using the factor contribution rate method, with a calculation window of 15 acquisition cycles. The influence coefficient for topology factors was 0.35, for valve-specific factors 0.4, and for state parameter factors 0.25. Specific parameters, influence coefficients, and influence trends for each factor were recorded and integrated to form multi-valve linkage burst pressure influencing factor data, clarifying the degree and law of influence of each factor on burst pressure release.

[0055] Step S42: Perform heterogeneous combination feature analysis of multi-valve linkage based on the multi-valve linkage control topology data of the battery system, and generate heterogeneous combination feature data of multi-valve linkage. In this embodiment of the invention, heterogeneous combination feature analysis of multi-valve linkage is carried out based on the multi-valve linkage control topology data of the battery system. The pressure relief valve execution nodes in the multi-valve linkage control topology data include two top pressure relief valves (F1, F2, model XY-01, flow area 150mm²) and four side wall pressure relief valves (F3-F6, model XY-02, flow area 120mm²). The linkage logic clearly defines the valve combinations and opening sequence under different risk scenarios. The analysis adopts a heterogeneous feature extraction method, dividing the heterogeneous combination dimensions according to valve type, installation position, and structural parameters. The first dimension is valve type heterogeneity, divided into top and side wall pressure relief valve combinations; the second dimension is installation position heterogeneity, divided into middle, tail, and side area valve combinations; the third dimension is structural parameter heterogeneity, divided into valve combinations with different flow areas and response times. For each heterogeneous combination dimension, the core features of the valves within the combination are extracted. In type-based heterogeneous combinations, the flow area difference between the top and sidewall valve combinations is 30 mm², and the response time difference is 5 ms. In position-based heterogeneous combinations, the overlap of coverage area between the middle and side valve combinations is 30%, and the overlap between the tail and side valve combinations is 25%. In structural parameter-based heterogeneous combinations, the pressure relief efficiency difference between valve combinations with a flow area of ​​150 mm² and 120 mm² is 0.33 mm² / ms. Simultaneously, the synergistic effect of each heterogeneous combination is analyzed. The synergistic effect is calculated as the ratio of the total pressure relief efficiency of the combination to the sum of the pressure relief efficiency of a single valve. A ratio greater than 0.9 is considered a good synergistic effect. The dimensions, core features, synergistic effects, and applicable risk scenarios of each heterogeneous combination are recorded and integrated to form multi-valve linkage heterogeneous combination feature data, clarifying the characteristics and synergistic effects of different heterogeneous combinations.

[0056] Step S43: Perform nonlinear time-series trend analysis of the influencing factors of multi-valve linkage explosion pressure using multi-valve linkage explosion pressure influencing factor data and multi-valve linkage heterogeneous combination characteristic data, and generate multi-valve linkage explosion pressure influencing factor trend data. In this embodiment of the invention, nonlinear time-series trend analysis of the influencing factors of multi-valve linkage explosion pressure is carried out by using multi-valve linkage explosion pressure influencing factor data and multi-valve linkage heterogeneous combination characteristic data. A nonlinear time-series analysis algorithm is adopted, with one acquisition cycle as the time step and the analysis time is set to 100 acquisition cycles, focusing on the time-series change law of the core explosion pressure influencing factors (internal pressure fluctuation amplitude, temperature change rate, valve response time, and heterogeneous combination synergistic effect). The time-series trend analysis of internal pressure fluctuation amplitude employed a nonlinear fitting algorithm. The fitting equation was a quadratic polynomial with a fitting coefficient set at 0.85, tracking the trend of internal pressure change from an initial value of 0.1 MPa over time. In high-risk scenarios, the internal pressure increased by 0.1 MPa every 10 acquisition cycles. The time-series trend analysis of temperature change rate used a slope calculation method. In high-risk scenarios, the slope of the temperature change rate was 0.5℃ / s², gradually increasing over time. The valve response time remained stable, consistently between 45-50 ms. The time-series trend analysis of heterogeneous combination synergy effect used a difference calculation method. In high-risk scenarios, the synergy effect increased by 0.05 every 15 acquisition cycles. Simultaneously, the nonlinear correlations between various influencing factors were analyzed. The nonlinear correlation coefficient between internal pressure fluctuation amplitude and temperature change rate was set at 0.75, and the nonlinear correlation coefficient between internal pressure fluctuation amplitude and heterogeneous combination synergy effect was set at 0.8. The time-series change curves, slopes, correlation coefficients, and trend characteristics of each influencing factor were recorded and integrated to form trend data of multi-valve linkage explosion pressure influencing factors, capturing the dynamic change patterns of explosion pressure influencing factors.

[0057] Step S44: Perform dynamic prediction processing of explosion pressure evolution based on the trend data of multi-valve linkage explosion pressure influencing factors to generate dynamic prediction data of explosion pressure evolution; In this embodiment of the invention, dynamic prediction processing of explosion pressure evolution is carried out based on the trend data of multi-valve linkage explosion pressure influencing factors. A dynamic prediction algorithm is employed, using the time-series trend data of explosion pressure influencing factors as a basis. The prediction step size is set to 5 acquisition cycles, and the prediction duration is 50 acquisition cycles, focusing on the prediction of three core explosion pressure evolution parameters: internal pressure, temperature, and pressure relief efficiency. Internal pressure evolution prediction uses a time-series recursive algorithm, calculating the internal pressure value for each prediction step based on the time-series trend of internal pressure fluctuation amplitude. In high-risk scenarios, the internal pressure is predicted to increase from 0.6 MPa to 1.0 MPa within 50 acquisition cycles. Temperature evolution prediction uses a trend extension algorithm, extending the calculation of temperature values ​​within the prediction duration based on the time-series slope of the temperature change rate. The temperature is predicted to increase from 80℃ to 120℃ within 50 acquisition cycles. Pressure relief efficiency evolution prediction combines the trend of heterogeneous combination synergistic effects to calculate the total pressure relief efficiency for each prediction step. In high-risk scenarios, the total pressure relief efficiency increases from 3 mm² / ms to 3.8 mm² / ms. Simultaneously, it predicts the risk level changes during the explosion pressure evolution process, maintains a high risk level within 50 collection cycles, expands the coverage from 50% to 70%, records the core evolution parameters, risk level and coverage of each prediction step, and integrates them to form dynamic prediction data of explosion pressure evolution. It predicts the evolution trend and change pattern of explosion pressure in advance, provides a predictive basis for the design of subsequent explosion pressure release control strategies, and ensures that the control strategies can adapt to the changes in explosion pressure evolution in advance.

[0058] Step S45: Design the explosion pressure release control strategy for the multi-valve linkage of the battery system using the multi-valve linkage control topology data and the dynamic prediction data of explosion pressure evolution, and generate the intelligent control strategy for explosion pressure release of the multi-valve linkage of the battery system.

[0059] In this embodiment of the invention, a multi-valve linkage control strategy for releasing burst pressure in a battery system is designed using topology data of the battery system's multi-valve linkage control and dynamic prediction data of burst pressure evolution. A hierarchical strategy design method is adopted, combining burst pressure evolution parameters and risk levels for different prediction periods to design corresponding control strategies. The control strategy is divided into three control stages. The first stage is the initial prediction phase (1-15 acquisition cycles), with internal pressure ≤0.7MPa and temperature ≤90℃. Low-intensity linkage control is used, opening two side-wall pressure relief valves (F3, F4) at 50% opening, maintaining a pressure relief efficiency of 2.67mm² / ms, and setting the trigger threshold based on an internal pressure of 0.4MPa and a temperature change rate of 5℃ / s. The second stage is the middle prediction phase (16-35 acquisition cycles), with internal pressure 0.7-0.9MPa and temperature 90-110℃. Medium-intensity linkage control is used, opening two side-wall pressure relief valves (F3, F4, F5, F6, F7, F8, F9 ... 4) One top pressure relief valve (F1) is opened at 75%, maintaining a pressure relief efficiency of 3.2 mm² / ms. The trigger threshold is set based on an internal pressure of 0.5 MPa and a temperature change rate of 6℃ / s. The third stage is the later stage of prediction (36-50 acquisition cycles), with an internal pressure ≥0.9 MPa and a temperature ≥110℃. High-intensity linkage control is adopted, opening four side wall pressure relief valves (F3-F6) and two top pressure relief valves (F1, F2) at 100% opening, maintaining a pressure relief efficiency of 3.8 mm² / ms. The trigger threshold is set based on an internal pressure of 0.6 MPa and a temperature change rate of 7℃ / s. At the same time, a strategy adjustment logic is designed. When the difference between the predicted explosion pressure evolution parameters and the actual acquired parameters exceeds 10%, the valve opening degree and number of valves are adjusted, with an adjustment step size of 25%, to ensure that the control strategy can dynamically adapt to the changes in explosion pressure evolution. By integrating the strategy parameters, trigger thresholds, valve opening schemes, and adjustment logic of the three control stages, a multi-valve linkage intelligent control strategy for burst pressure release in the battery system is formed. This strategy enables precise and dynamic control of burst pressure release, effectively prevents thermal runaway risks, ensures the safety of the battery system and its surroundings, and meets the core requirements of multi-valve linkage burst pressure release.

[0060] Furthermore, step S45 includes the following steps: Step S451: Perform action space analysis of multi-valve collaborative control strategy based on the multi-valve linkage control topology data of the battery system, and generate action space data of multi-valve collaborative control strategy; In this embodiment of the invention, action space analysis of the multi-valve collaborative control strategy is carried out based on the multi-valve linkage control topology data of the battery system. The multi-valve linkage control topology data includes six pressure relief valve execution nodes (identified as F1-F6, where F1 and F2 are top pressure relief valves, model XY-01, with a flow area of ​​150mm²; F3-F6 are side wall pressure relief valves, model XY-02, with a flow area of ​​120mm²), state nodes, connection weights, and linkage logic. The connection weights are set according to internal pressure 0.5, temperature 0.3, and strain 0.2. The linkage logic clarifies the valve opening sequence under different risk scenarios. The analysis adopts the action space modeling method, taking the opening action of the pressure relief valve as the core action unit. The action space is divided into three action dimensions: the number of valves opened, the opening degree, and the opening sequence. Each dimension has a clearly defined range of action parameters. The number of valves opened is categorized into 1-6 valves, corresponding to pressure relief requirements in different risk scenarios. The opening degree is divided into four levels: 25%, 50%, 75%, and 100%, set according to pressure relief intensity. The opening sequence is categorized into three modes based on installation location: side-wall priority, top priority, and synchronous opening. In side-wall priority mode, the top valve opens at a 20ms interval; in top priority mode, the side-wall valve opens at a 15ms interval; and in synchronous opening mode, there is no time interval. The parameters for each action dimension are quantified and defined, clarifying the execution node, action parameters, and triggering conditions for each action. The feasibility of different action combinations is analyzed, eliminating invalid actions that do not match the number of valves opened or the pressure relief requirements, or the opening degree and the burst pressure intensity. The parameters, execution logic, and applicable scenarios of all valid actions are recorded, integrating them to form the action space data for the multi-valve collaborative control strategy, clearly presenting all executable actions and parameter ranges for multi-valve collaborative control.

[0061] Step S452: Based on the action space data of the multi-valve collaborative control strategy, perform an action space explosion pressure release benefit analysis on the dynamic prediction data of explosion pressure evolution, and generate action space explosion pressure release benefit data; In this embodiment of the invention, the explosion pressure release benefit analysis is performed on the dynamic prediction data of explosion pressure evolution based on the action space data of the multi-valve collaborative control strategy. The dynamic prediction data of explosion pressure evolution includes the evolution parameters of internal pressure, temperature, and pressure relief efficiency within 50 acquisition cycles. In high-risk scenarios, the internal pressure increases from 0.6 MPa to 1.0 MPa, the temperature increases from 80°C to 120°C, and the total pressure relief efficiency increases from 3 mm² / ms to 3.8 mm², expanding the coverage from 50% to 70%. The analysis adopts a benefit quantification algorithm, taking each effective action combination in the action space as the analysis unit, and calculates the explosion pressure release benefit of each action combination. The release benefit calculation formula is the product of the pressure relief efficiency increase value and the explosion pressure parameter decrease value, where the explosion pressure parameter decrease value is the sum of the internal pressure decrease amplitude and the temperature decrease amplitude. A benefit threshold of 1.2 was set, and the benefit of each action combination was calculated. For example, the action combination of opening two side wall valves (F3, F4) at 50% opening and prioritizing side wall opening resulted in a pressure relief efficiency improvement of 0.27 mm² / ms, an internal pressure drop of 0.1 MPa, and a temperature drop of 5℃, with a release benefit of 0.27 × (0.1 + 5) = 1.37. The action combination of opening four side wall valves (F3-F6) and two top valves (F1, F2) at 100% opening and opening simultaneously resulted in a pressure relief efficiency improvement of 0.8 mm² / ms, an internal pressure drop of 0.4 MPa, and a temperature drop of 40℃, with a release benefit of 0.8 × (0.4 + 40) = 32.32. The release benefit value, pressure relief efficiency improvement, and pressure explosion parameter drop of each action combination were recorded. High-efficiency action combinations with benefits exceeding the threshold were marked. The benefit data and related parameters of all action combinations were integrated to form action space pressure explosion release benefit data.

[0062] Step S453: Analyze the optimization requirements of the battery system's explosion pressure evolution using dynamic prediction data of explosion pressure evolution, and generate optimization requirement data for the battery system's explosion pressure evolution. In this embodiment of the invention, optimization target analysis of battery system explosion pressure evolution is conducted using dynamic prediction data of explosion pressure evolution. The dynamic prediction data clearly shows that a high-risk level is maintained within 50 acquisition cycles, with internal pressure and temperature continuously rising, pressure relief efficiency gradually improving, and the coverage area continuously expanding. The analysis employs a target decomposition method, dividing the optimization target of explosion pressure evolution into two levels: core targets and secondary targets. The core target is to control the internal pressure to not exceed 1.0 MPa and the temperature to not exceed 120°C, ensuring that the battery casing does not rupture. The secondary targets are to improve pressure relief efficiency, reduce the risk coverage area, and decrease the thermal runaway propagation speed. Quantitative standards are set for the core targets: internal pressure control target is ≤1.0 MPa, temperature control target is ≤120°C, and risk coverage control target is ≤70%. Quantitative standards are set for the secondary targets: pressure relief efficiency improvement target is ≥3.8 mm² / ms, and thermal runaway propagation speed control target is ≤0.1 m / s. Simultaneously, the optimization requirements for different prediction periods were analyzed. In the initial prediction phase (1-15 data acquisition cycles), the optimization requirement was to suppress the rapid rise in internal pressure, focusing on improving the pressure relief efficiency to above 2.8 mm² / ms. In the mid-prediction phase (16-35 data acquisition cycles), the optimization requirement was to control the rate of temperature rise, keeping the temperature change rate below 6℃ / s. In the late prediction phase (36-50 data acquisition cycles), the optimization requirement was to maintain high pressure relief efficiency, ensuring that the internal pressure did not exceed 1.0 MPa. The optimization objectives, quantification standards, and optimization priorities for each level were recorded and integrated to form the optimization requirement target data for battery system explosion pressure evolution, clarifying the optimization direction and quantification requirements for explosion pressure evolution.

[0063] Step S454: Design a multi-valve linkage intelligent control strategy for the release of burst pressure in the battery system based on the burst pressure release benefit data of the action space and the target data of the burst pressure evolution optimization requirements of the battery system.

[0064] In this embodiment of the invention, a multi-valve linkage intelligent control strategy for the release of burst pressure in the battery system is designed based on the burst pressure release benefit data of the action space and the target data of the battery system burst pressure evolution optimization requirements. A target-benefit matching method is adopted to accurately match the optimization requirements with efficient action combinations. A hierarchical control strategy is designed based on the optimization focus of different prediction periods. In the initial prediction phase (1-15 acquisition cycles), the optimization requirement is to suppress the rapid rise of internal pressure. Action combinations with release benefits of 1.37-5.0 are matched, and low-intensity linkage control is adopted. Two side-wall pressure relief valves (F3, F4) are opened at 50% opening, with priority given to side-wall valves. The trigger threshold is set based on an internal pressure of 0.4 MPa and a temperature change rate of 5℃ / s to ensure that the pressure relief efficiency is increased to above 2.8 mm² / ms, suppressing the rate of internal pressure rise. During the mid-term forecast (16-35 acquisition cycles), the optimization requirement is to control the rate of temperature rise, match the action combination with a release benefit of 5.0-20.0, adopt medium-intensity linkage control, open two side wall pressure relief valves (F3, F4) and one top pressure relief valve (F1), with an opening degree of 75%, and the side walls are opened first (the top valve is opened at 20ms intervals). The trigger threshold is set according to the internal pressure of 0.5MPa and the temperature change rate of 6℃ / s, controlling the temperature change rate below 6℃ / s, while improving the pressure relief efficiency to above 3.2mm² / ms. In the later stages of the forecast (36-50 data acquisition cycles), the optimization requirements are to maintain high pressure relief efficiency, control the internal pressure to no more than 1.0 MPa, and match action combinations with a release benefit ≥20.0. High-intensity linkage control is employed, opening four side-wall pressure relief valves (F3-F6) and two top pressure relief valves (F1, F2) at 100% opening, simultaneously. Trigger thresholds are set based on an internal pressure of 0.6 MPa and a temperature change rate of 7℃ / s, ensuring that the pressure relief efficiency is increased to above 3.8 mm² / ms, the internal pressure is controlled within 1.0 MPa, and the temperature is controlled within 120℃. Simultaneously, a closed-loop adjustment logic is designed. Every 5 data acquisition cycles, the difference between the actual burst pressure parameters and the optimization target is compared. When the difference exceeds 10%, the opening degree and number of actions in the action combination are adjusted, with an adjustment step size of 25%, ensuring that the control strategy can dynamically adapt to changes in burst pressure and achieve the optimization target. By integrating the action combinations, trigger thresholds, activation parameters, and adjustment logic of the hierarchical control strategy, a multi-valve linkage intelligent control strategy for burst pressure release in the battery system is formed. This strategy enables precise and dynamic control of burst pressure release, effectively prevents thermal runaway risks, ensures the safety of the battery system and its surroundings, and meets the core requirements of multi-valve linkage burst pressure release.

[0065] This specification provides a multi-valve linkage burst pressure release system for a battery system, used to execute the multi-valve linkage burst pressure release method for a battery system as described above. The multi-valve linkage burst pressure release system for a battery system includes: The battery system operation status monitoring and analysis module is used to monitor and process the battery system operation status using a multi-type sensor array deployed inside the battery module and casing, and generate battery system operation status monitoring data; based on the battery system operation status monitoring data, it performs multi-dimensional time-series state characteristic analysis of the battery system, and generates multi-dimensional time-series state characteristic data of the battery system. The thermal runaway propagation risk characteristic analysis module is used to perform thermal runaway propagation risk characteristic analysis on battery system based on multi-dimensional time-series state characteristic data of battery system, and generate thermal runaway propagation risk characteristic data of battery system; The battery system multi-valve linkage control topology analysis module is used to acquire the structural configuration data of the battery housing pressure relief valve; and to perform multi-valve linkage control topology analysis on the multi-dimensional time-series state characteristic data of the battery system using the structural configuration data of the battery housing pressure relief valve and the thermal runaway propagation risk characteristic data of the battery system, thereby generating multi-valve linkage control topology data of the battery system. The intelligent control module for multi-valve linkage pressure release in battery systems is used to design pressure release control strategies for multi-valve linkage in battery systems based on the multi-valve linkage control topology data of the battery system, and to generate intelligent control strategies for multi-valve linkage pressure release in battery systems.

[0066] Therefore, the embodiments should be considered as exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of the equivalents of the application are intended to be included within the invention.

[0067] The above description is merely a specific embodiment of the present invention, enabling those skilled in the art to understand or implement the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the present invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features of the invention herein.

Claims

1. A method for releasing burst pressure in a battery system using multiple valves, characterized in that, Includes the following steps: Step S1: Utilize a multi-type sensor array deployed inside the battery module and casing to monitor and process the battery system's operating status, generating battery system operating status monitoring data; based on the battery system operating status monitoring data, perform multi-dimensional time-series state characteristic analysis of the battery system, generating multi-dimensional time-series state characteristic data of the battery system. Step S2: Analyze the thermal runaway propagation risk characteristics of the battery system based on the multi-dimensional time-series state characteristic data of the battery system, and generate thermal runaway propagation risk characteristic data of the battery system; Step S3: Obtain the structural configuration data of the battery housing pressure relief valve; perform multi-valve linkage control topology analysis on the multi-dimensional time-series state characteristic data of the battery system using the structural configuration data of the battery housing pressure relief valve and the thermal runaway propagation risk characteristic data of the battery system, and generate multi-valve linkage control topology data of the battery system. Step S4: Design the burst pressure release control strategy for the multi-valve linkage of the battery system based on the multi-valve linkage control topology data of the battery system, and generate the intelligent control strategy for burst pressure release of the multi-valve linkage of the battery system.

2. The multi-valve linkage explosion pressure release method for a battery system according to claim 1, characterized in that, Step S1 includes the following steps: Step S11: Use a multi-type sensor array deployed inside the battery module and casing to monitor and process the battery system's operating status, and generate battery system operating status monitoring data; Step S12: Perform multi-source response analysis of battery system operating status based on battery system operating status monitoring data to generate multi-source response data of battery system operating status; Step S13: Perform multi-physics coupling deconstruction processing on the battery system based on the multi-source response data of the battery system operating status to generate multi-physics coupling deconstruction data of the battery system; Step S14: Perform battery system response decoupling characteristic analysis based on the multi-physics coupling deconstruction data of the battery system to generate battery system response decoupling characteristic data; Step S15: Perform spatial energy and matter migration mapping feature analysis using battery system response decoupling feature data to generate battery system energy-matter migration mapping feature data; Step S16: Based on the battery system response decoupling feature data and the battery system energy-mass migration mapping feature data, perform adaptive dimensionality reduction and resampling processing of the battery system response migration sensitive region to generate dimensionality reduction response decoupling feature data of the battery system. Step S17: Perform multi-dimensional time-series state feature analysis on the battery system distribution dimensionality reduction of the battery system response decoupling feature data and the battery system energy-mass migration mapping feature data through the battery system dimensionality reduction response decoupling feature data, and generate multi-dimensional time-series state feature data of the battery system.

3. The multi-valve linkage explosion pressure release method for a battery system according to claim 2, characterized in that, The multi-source response data of the battery system operating status mentioned in step S12 includes temperature dynamic response data, gas concentration change response data, internal pressure transient fluctuation response data, voltage and current disturbance response data, and casing strain response data.

4. The multi-valve linkage explosion pressure release method for a battery system according to claim 2, characterized in that, Step S14 includes the following steps: Step S141: Perform multi-physics decoupling response gradient distribution analysis on the multi-physics coupling deconstruction data of the battery system to generate multi-physics decoupling response gradient distribution data; Step S142: Perform dominant feature analysis of local response of multiphysics decoupling based on the gradient distribution data of multiphysics decoupling response, and generate dominant feature data of local response of multiphysics decoupling response; Step S143: Perform correlation analysis on the multi-physics decoupling based on the multi-physics coupling deconstruction data of the battery system, and generate multi-physics decoupling correlation data; Step S144: Perform battery system response decoupling feature analysis using multi-physics decoupling local response dominant feature data and multi-physics decoupling correlation data to generate battery system response decoupling feature data.

5. The multi-valve linkage explosion pressure release method for a battery system according to claim 1, characterized in that, Step S2 includes the following steps: Step S21: Perform multi-layer heterogeneous characteristic analysis of the battery system based on the multi-dimensional time-series state characteristic data of the battery system, generate multi-layer heterogeneous characteristic data of the battery system, and perform multi-layer heterogeneous structural node modeling processing through the multi-layer heterogeneous characteristic data of the battery system to generate multi-layer heterogeneous structural node data. Step S22: Analyze the multi-mechanism propagation relationship of the battery system based on the multi-layer heterogeneous structure node data, generate multi-mechanism propagation relationship data of the battery system, and analyze the multi-mechanism propagation driving factors through the multi-mechanism propagation relationship data of the battery system to generate multi-mechanism propagation driving factor data. Step S23: Analyze the propagation path distribution characteristics based on the multi-layer heterogeneous structure node data and the multi-mechanism propagation driving factor data to generate propagation path distribution characteristic data; Step S24: Analyze the risk clustering characteristics of the transmission path based on the distribution characteristic data of the transmission path, and generate risk clustering characteristic data of the transmission path; Step S25: Analyze the propagation risk characteristics of the battery system through the propagation path risk clustering feature data, and generate the thermal runaway propagation risk feature data of the battery system.

6. The multi-valve linkage explosion pressure release method for a battery system according to claim 1, characterized in that, Step S3 includes the following steps: Step S31: Obtain the structural configuration data of the battery housing pressure relief valve; Step S32: Based on the battery casing pressure relief valve structure configuration data and battery system thermal runaway propagation risk characteristic data, perform thermal runaway risk and valve action mapping relationship analysis to generate thermal runaway risk-valve action mapping relationship data; Step S33: Based on the thermal runaway risk-valve action mapping relationship data, perform multi-valve linkage candidate combination analysis for thermal runaway risk to generate multi-valve linkage candidate combination data; perform constraint optimization solution on the multi-valve linkage candidate combination data to generate multi-valve linkage combination optimization data; Step S34: Based on the multi-valve linkage combination optimization data, perform multi-valve linkage control topology analysis on the multi-dimensional time-series state characteristic data of the battery system to generate multi-valve linkage control topology data of the battery system.

7. The multi-valve linkage explosion pressure release method for a battery system according to claim 6, characterized in that, Step S32 includes the following steps: Step S321: Based on the battery housing pressure relief valve structure configuration data, perform pressure relief valve execution node modeling to generate battery system pressure relief valve execution node data; Step S322: Perform a function correlation feature analysis on the pressure relief valve execution node data of the battery system to generate function correlation feature data of the pressure relief valve execution node; Step S323: Perform relational abstraction processing on the thermal runaway propagation risk characteristic data of the battery system to generate abstract characteristic data of thermal runaway propagation risk; Step S324: Map the function-related feature data of the pressure relief valve to the abstract feature data of thermal runaway propagation risk to perform thermal runaway risk and valve function mapping relationship analysis, and generate thermal runaway risk-valve function mapping relationship data.

8. The multi-valve linkage explosion pressure release method for a battery system according to claim 1, characterized in that, Step S4 includes the following steps: Step S41: Analyze the factors affecting the burst pressure of the multi-valve linkage control topology based on the multi-valve linkage control topology data of the battery system, and generate multi-valve linkage burst pressure influencing factor data; Step S42: Perform heterogeneous combination feature analysis of multi-valve linkage based on the multi-valve linkage control topology data of the battery system, and generate heterogeneous combination feature data of multi-valve linkage. Step S43: Perform nonlinear time-series trend analysis of the influencing factors of multi-valve linkage explosion pressure using multi-valve linkage explosion pressure influencing factor data and multi-valve linkage heterogeneous combination characteristic data, and generate multi-valve linkage explosion pressure influencing factor trend data. Step S44: Perform dynamic prediction processing of explosion pressure evolution based on the trend data of multi-valve linkage explosion pressure influencing factors to generate dynamic prediction data of explosion pressure evolution; Step S45: Design the explosion pressure release control strategy for the multi-valve linkage of the battery system using the multi-valve linkage control topology data and the dynamic prediction data of explosion pressure evolution, and generate the intelligent control strategy for explosion pressure release of the multi-valve linkage of the battery system.

9. The multi-valve linkage explosion pressure release method for a battery system according to claim 8, characterized in that, Step S45 includes the following steps: Step S451: Perform action space analysis of multi-valve collaborative control strategy based on the multi-valve linkage control topology data of the battery system, and generate action space data of multi-valve collaborative control strategy; Step S452: Based on the action space data of the multi-valve collaborative control strategy, perform an action space explosion pressure release benefit analysis on the dynamic prediction data of explosion pressure evolution, and generate action space explosion pressure release benefit data; Step S453: Analyze the optimization requirements of the battery system's explosion pressure evolution using dynamic prediction data of explosion pressure evolution, and generate optimization requirement data for the battery system's explosion pressure evolution. Step S454: Design a multi-valve linkage intelligent control strategy for the release of burst pressure in the battery system based on the burst pressure release benefit data of the action space and the target data of the burst pressure evolution optimization requirements of the battery system.

10. A multi-valve linkage pressure release system for a battery system, characterized in that, For performing the battery system multi-valve linkage burst pressure release method as described in claim 1, the battery system multi-valve linkage burst pressure release system includes: The battery system operation status monitoring and analysis module is used to monitor and process the battery system operation status using a multi-type sensor array deployed inside the battery module and casing, and generate battery system operation status monitoring data; based on the battery system operation status monitoring data, it performs multi-dimensional time-series state characteristic analysis of the battery system, and generates multi-dimensional time-series state characteristic data of the battery system. The thermal runaway propagation risk characteristic analysis module is used to perform thermal runaway propagation risk characteristic analysis on battery system based on multi-dimensional time-series state characteristic data of battery system, and generate thermal runaway propagation risk characteristic data of battery system; The battery system multi-valve linkage control topology analysis module is used to acquire the structural configuration data of the battery housing pressure relief valve; and to perform multi-valve linkage control topology analysis on the multi-dimensional time-series state characteristic data of the battery system using the structural configuration data of the battery housing pressure relief valve and the thermal runaway propagation risk characteristic data of the battery system, thereby generating multi-valve linkage control topology data of the battery system. The intelligent control module for multi-valve linkage pressure release in battery systems is used to design pressure release control strategies for multi-valve linkage in battery systems based on the multi-valve linkage control topology data of the battery system, and to generate intelligent control strategies for multi-valve linkage pressure release in battery systems.

Images