Energy storage battery thermal runaway damage diffusion early warning method based on dynamic thermal model

Abstract

The invention discloses an energy storage battery thermal runaway damage diffusion early warning method based on a dynamic thermal model, and relates to the technical field of battery safety monitoring, and the early warning method comprises the specific steps: multi-source parameter collection: arranging a plurality of sensors to collect data, and preprocessing and storing the data; thermal stress modeling is conducted, units are divided through a finite element method to calculate thermal stress, and structural damage is judged in combination with a damage threshold value; multi-physics field cooperative calibration is carried out, and characteristic quantity is acquired and substituted into a formula to correct thermal stress; dynamic thermal model construction: introducing thermal stress influence to construct a model containing a damage diffusion term; thermal runaway early warning judgment is carried out, a multi-dimensional index system is constructed, a risk value is calculated, and early warning is triggered; through a multi-source parameter acquisition technology, in combination with temperature, current, voltage, ultrasonic wave and broadband impedance multi-dimensional sensor data, and assisted by external condition monitoring of environment temperature and humidity, heat dissipation system flow and rotation speed, comprehensive perception of the working state of the battery is realized, and richness and accuracy of data acquisition are improved.

Description

Technical Field

[0001] This invention relates to the field of battery safety monitoring technology, specifically to a method for early warning of thermal runaway damage propagation in energy storage batteries based on a dynamic thermal model. Background Technology

[0002] With the rapid development of new energy vehicles and energy storage systems, lithium-ion batteries have become the mainstream choice due to their high energy density and long cycle life. However, batteries generate a lot of heat during charging and discharging. If heat dissipation is not timely, the internal temperature of the battery will rise sharply, which may lead to battery thermal runaway, causing fire or even explosion. Battery thermal runaway not only threatens the safety of passengers and equipment, but also causes huge economic losses and environmental impacts. Therefore, it is particularly important to develop an efficient and accurate early warning method for the spread of battery thermal runaway damage.

[0003] Application No. 202311814165.8, a method for early warning of thermal runaway in battery cabinets based on internal battery temperature estimation, improves the accuracy of early warning by estimating the internal battery temperature through a thermodynamic model. However, it mainly relies on the battery charging and discharging current, voltage, and internal resistance parameters to construct the model, failing to fully consider the impact of thermal stress on battery structural damage and making it difficult to quantify the risk of irreversible material deformation caused by temperature gradients. In terms of data acquisition, it only uses traditional temperature and current sensors, lacking the coordinated monitoring of ultrasonic and broadband impedance multi-physics field characteristics, and cannot achieve the results of sound velocity-structural deformation and impedance analysis. - The mapping relationship of contact resistance is used to calibrate the thermal model deviation in real time; although a thermal runaway risk index is introduced into the early warning index system, the cumulative effect of thermal history is not described by fractional derivatives, which is insufficient for the dynamic adaptability of battery aging state (SOH) and operating conditions (SOC, charge and discharge current). Moreover, a multi-dimensional hierarchical early warning strategy is not constructed, making it difficult to achieve precise intervention at different stages of thermal runaway. In addition, the method does not involve finite element discretization modeling and damage diffusion term analysis, which limits the characterization of the internal thermo-mechanical coupling behavior of the battery. The robustness and operating condition adaptability of the early warning method need to be further improved. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of existing technologies and provide a method for early warning of thermal runaway damage propagation in energy storage batteries based on a dynamic thermal model. By acquiring multi-source parameters and co-calibrating with multi-physics fields, a dynamic thermal model containing damage propagation terms is constructed. This method uses the characteristic quantities of ultrasound and impedance to correct thermal stress deviations in real time, and combines fractional derivatives to describe the cumulative effect of thermal history, dynamically adapting to battery aging and changes in operating conditions. The early warning system calculates the risk value through a fuzzy-entropy weight decision function and triggers a graded response strategy, realizing accurate early warning and safety intervention for thermal runaway.

[0005] To solve the above-mentioned technical problems, this invention provides the following technical solution: a method for early warning of thermal runaway damage propagation in energy storage batteries based on a dynamic thermal model, the specific steps of which are as follows:

[0006] S100, multi-source parameter acquisition: Multiple sensors are arranged in the battery and module to collect data on electrical parameters, temperature environment and heat dissipation system. Battery material characteristic parameters are obtained from the material database. At the same time, ultrasonic sensors and broadband impedance testers are used to collect multi-physics field characteristic quantities. The collected data are preprocessed, standardized and stored in the database.

[0007] S200, Thermal Stress Modeling: Based on thermoelasticity theory, the battery is divided into tiny units using the finite element discretization method. A thermal stress model is constructed to calculate the thermal stress of each unit. The damage threshold is defined by combining the critical value of irreversible deformation of thermal stress of battery material. The influence of battery usage stage and operating conditions on the threshold is considered to determine whether thermal stress causes irreversible damage to the battery structure.

[0008] S300, Multi-physics Field Collaborative Calibration: Simultaneously acquires ultrasonic propagation speed, AC impedance phase angle, battery structure deformation, and actual contact resistance values; analyzes the mapping relationship between ultrasonic speed and structural deformation, and impedance phase angle and contact resistance; transforms the changes in characteristic quantities into changes in battery structure deformation and contact resistance; substitutes these values ​​into the multi-field collaborative topology mapping calibration formula; and corrects the thermal stress in real time.

[0009] S400, Dynamic Thermal Model Construction: Based on the traditional heat generation rate model, the influence of thermal stress on the structure and thermal properties is introduced, and a dynamic thermal model including damage diffusion term is constructed by combining the calibrated parameters.

[0010] S500, thermal runaway early warning judgment: Construct a multi-dimensional early warning indicator system that includes temperature, thermal stress, and thermal accumulation trend. Calculate the risk value through a fuzzy-entropy weighted early warning decision function and trigger the corresponding level of early warning according to the preset threshold.

[0011] Furthermore, in S100, the multi-source parameter sampling is concentrated in sensors arranged in the battery and module, namely:

[0012] Temperature sensor: installed on the surface of individual battery cells and near the heat dissipation channels of the battery module;

[0013] Current sensor: installed in series in the current loop of the battery module;

[0014] Voltage sensor: installed in parallel at both ends of a battery cell or at the output end of a battery module;

[0015] Ultrasonic sensors: installed between the cells of the battery module and inside the module housing, used to transmit and receive high-frequency ultrasonic signals;

[0016] Wideband impedance tester: connected to the electrode output terminal of the battery module;

[0017] Ambient temperature sensor: located on the outside of the battery module;

[0018] Humidity sensor: installed in the external environment of the battery module;

[0019] Flow sensor: Installed in the coolant piping of the battery liquid cooling system;

[0020] Speed ​​sensor: Installed at the cooling fan motor of the battery air-cooling system.

[0021] Furthermore, in S200, during the process of dividing the battery into micro-units in thermal stress modeling, the specific steps for dividing the battery into micro-units using the finite element discretization method are as follows:

[0022] (1) Using 3D modeling software, a 3D solid model of the battery containing each component is accurately created based on the actual size and component parameters of the battery.

[0023] (2) Select finite element analysis software and appropriate element type, and in combination with analysis requirements, set the element size of different areas of the battery to the millimeter level, and determine the discretization strategy.

[0024] (3) Import the 3D model into the software to perform mesh generation, generate a finite element mesh composed of tiny units, and check and optimize the mesh quality;

[0025] (4) Assign thermo-mechanical property parameters to each micro-unit, and define the heat exchange and mechanical constraint conditions of the unit;

[0026] (5) Set the thermal stress solver parameters in the software, load the collected temperature data as the load, and complete the preprocessing for thermal stress calculation.

[0027] Furthermore, in S200, the construction of the thermal stress model in thermal stress modeling, the calculation formula of the thermal stress model is as follows: Where, σ th Thermal stress represents the stress generated inside the battery due to temperature changes. E is the elastic modulus, the material's ability to resist elastic deformation, obtained from a material database. r is the coefficient of thermal expansion, the material's expansion characteristics with temperature changes, obtained through material testing. ΔT is the temperature difference between the current and previous moments. ν is Poisson's ratio, the ratio of the material's lateral strain to its longitudinal strain, determined by material properties. x, y, and z are spatial coordinates used to locate the positions of the tiny cells inside the battery. It is the temperature gradient, calculated using the central difference method.

[0028] Furthermore, in S200, the calculation formula for the thermal stress damage threshold in thermal stress modeling is: σcr =σ cr0 ·λ(SOH)·μ(SOC,I), where σ cr It is the critical value for thermal stress damage, the stress threshold at which battery materials undergo irreversible deformation, σ cr0 The initial damage threshold is the baseline damage threshold for a new battery. λ(SOH) is the health state correction coefficient, reflecting the impact of battery aging on the threshold, fitted through cycle life experiments. μ(SOC,I) is the operating condition correction function, obtained through neural network training based on charge / discharge current I and SOC data. SOH represents the battery health state, characterizing the degree of battery capacity decay. SOC is the state of charge, the ratio of remaining battery capacity to rated capacity. I is the charge / discharge current, collected by a current sensor. When σ... th <0.8σ cr When the deformation is 0.8σ, it is determined to be elastic deformation, and the structure can recover; cr ≤σ th <σ cr When σ is triggered, a damage warning is issued; when σ is triggered, a damage warning is issued. th ≥σ cr At that time, it was determined to be an irreversible deformation.

[0029] Furthermore, in S300, the calculation formula for the multi-field cooperative topology mapping calibration formula in multi-physics cooperative calibration is as follows: Where Δσ th This is the thermal stress correction factor, used to correct the deviation of the thermal stress model in real time. ω is the multi-field calibration weight, determined through Bayesian optimization training. The optimization objective is to minimize the thermal stress prediction error. Δv ultra It is the change in ultrasonic velocity, the difference between the current sound velocity and the initial healthy state sound velocity, collected by an ultrasonic sensor, ΔZ. imped It is the impedance change, the difference between the current impedance and the initial healthy state impedance, acquired by a broadband impedance meter. β is the multi-field coupling exponent, which adjusts the weights of the ultrasonic and impedance signals, calibrated through thermal runaway experiments. It is the gradient operator, used to calculate the rate of change of spatial parameters, α. cal It is the calibrated coefficient of thermal expansion, combined with material parameters corrected using multiphysics data, E cal It is the calibrated elastic modulus, combined with material parameters corrected by multiphysics data, ΔS topo It is the entropy change of the topological network, the change in the disorder of the network constructed by multi-physics characteristic quantities, calculated using the information entropy formula, ΔS dam It is the damage entropy change, calculated by the damage entropy increase model in the thermal stress modeling process, where ΔT is the temperature difference between the current moment and the previous moment.

[0030] Furthermore, in S400, the calculation formula for the dynamic thermal model in the dynamic thermal model construction is as follows: in This is the Caputo fractional enthalpy change rate, describing the historical cumulative effect of enthalpy change. α∈(0,1) is the fractional order, ρ is the cell density, measured by the displacement method, t is the time variable, and c... p It is the constant-pressure heat capacity, obtained by measuring with a calorimeter. T represents the battery temperature. It is a divergence operator used to calculate the spatial distribution of heat flux density, k eff It is the effective thermal conductivity coefficient, including thermal conductivity parameters corrected for structural damage, I 2 R is the Joule heat yield, I is the current, R is the battery internal resistance, ξ is the damage diffusion coefficient, calibrated through thermal runaway experiments, characterizing the accelerating effect of damage on heat diffusion, and γ is the damage acceleration index, calibrated in conjunction with η in thermal stress modeling, reflecting the accelerating effect of damage on thermal runaway. It is the Laplace operator, used to calculate the spatial diffusion trend of damage entropy, S dam Damage entropy is calculated using the fractal dimension of the SEM image. ζ is the feedback coefficient, which converts the thermal stress correction into a proportionality coefficient for thermal generation compensation. Δσ th This is the thermal stress correction factor, used to correct deviations in the thermal stress model in real time, σ. cr It is the critical value for thermal stress damage, the stress threshold at which battery materials undergo irreversible deformation. Temperature gradient, where t is a time variable.

[0031] Furthermore, in S500, the calculation formula for the fuzzy-entropy weighted early warning decision in the thermal runaway early warning judgment is as follows: Where μ i It is the entropy weight coefficient, the indicator weight calculated using information entropy, R. risk This is the thermal runaway risk value; the higher the value, the higher the risk. 'n' represents the number of warning indicators, and 'H' represents the risk level. i H is the information entropy of the i-th indicator. max It is the historical maximum entropy value, Fuzzy(·) is the fuzzy membership function that maps continuous indicators to risk membership degrees, and uses a triangular fuzzy function to define the interval, T. i This is the i-th temperature index data, σ th,i This is the i-th thermal stress index data. It is the fractional-order enthalpy change rate, a heat accumulation trend index output by the dynamic thermal model.

[0032] Furthermore, in S500, the thermal runaway early warning judgment triggers corresponding level warnings based on preset thresholds:

[0033] When R risk <0.3 indicates low risk, and will only be recorded in the local logs of the battery management system, triggering a mild heat dissipation strategy;

[0034] When 0.3≤R risk If the value is less than 0.7, it indicates a medium risk. An early warning will be sent to the display terminal and remote monitoring platform, triggering an active temperature equalization strategy to limit the charging and discharging power to 80%.

[0035] When R risk A value of ≥0.7 indicates a high risk and will immediately trigger an audible and visual alarm, a push notification to the mobile app, an emergency alarm on the cloud platform, and coordinated safety intervention measures.

[0036] Compared with existing technologies, this early warning method for thermal runaway damage propagation in energy storage batteries based on a dynamic thermal model has the following advantages:

[0037] I. This invention utilizes multi-source parameter acquisition technology, combining data from multiple sensors including temperature, current, voltage, ultrasound, and wideband impedance sensors, supplemented by monitoring of environmental temperature and humidity, heat dissipation system flow rate, and rotational speed. This enables comprehensive perception of the battery's operating status, not only enhancing the richness and accuracy of data acquisition but also providing a solid data foundation for subsequent thermal stress modeling and dynamic thermal model construction. In particular, the introduction of a multi-physics field collaborative calibration mechanism effectively corrects deviations in the thermal stress model by analyzing the mapping relationship between ultrasonic propagation speed, AC impedance phase angle, battery structural deformation, and contact resistance in real time. This improves the real-time performance and accuracy of the early warning method. This cross-physics field comprehensive calibration method breaks through the limitations of traditional single-parameter monitoring, bringing new technical ideas to the field of battery safety early warning and enhancing adaptability and robustness to complex operating conditions.

[0038] Second, this invention incorporates the influence of thermal stress on battery structure and thermal characteristics into the traditional heat generation rate model and introduces a damage diffusion term, enabling the model to more realistically reflect the dynamic changes of the battery during thermal runaway. This improvement not only enhances the model's predictive ability but also provides a scientific basis for the accurate assessment of battery safety status. Furthermore, the thermal runaway early warning judgment system constructs multi-dimensional indicators including temperature, thermal stress, and thermal accumulation trends, and adopts a fuzzy-entropy weighted early warning decision function to achieve quantitative assessment and graded early warning of battery thermal runaway risk. Compared with single-indicator early warning, this comprehensive early warning method based on multi-dimensional indicators can more comprehensively capture subtle changes in battery safety status, detect potential thermal runaway risks in advance, and thus take more timely and effective intervention measures to ensure the safe operation of the battery system.

[0039] Other advantages, objectives and features of the invention will be set forth in part in the description which follows, and in part will be apparent to those skilled in the art from the following examination or study, or may be learned from the practice of the invention. Attached Figure Description

[0040] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0041] Figure 1 The flowchart shows a method for early warning of thermal runaway damage propagation in energy storage batteries based on a dynamic thermal model.

[0042] Figure 2 This is a framework diagram of an early warning method for thermal runaway damage propagation in energy storage batteries based on a dynamic thermal model. Detailed Implementation

[0043] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided below.

[0044] Example 1:

[0045] Early warning of thermal runaway of containerized batteries in large-scale energy storage power stations.

[0046] A certain new energy large-scale energy storage power station is equipped with multiple containerized energy storage units. Each container contains a large number of lithium-ion battery modules. During long-term charging and discharging, these batteries are at risk of thermal runaway. Once thermal runaway occurs, it may cause a fire or even an explosion, posing a serious threat to the safe operation of the power station. In this scenario, the present invention can realize accurate early warning of battery thermal runaway, ensuring the safe and stable operation of the power station.

[0047] S100, Multi-Source Parameter Acquisition: Temperature sensors are installed on the surface of each battery cell and near the heat dissipation channels of the battery module within each container to monitor temperature changes on the battery surface and in the heat dissipation channels in real time. Current sensors are connected in series in the current loop of the battery module, and voltage sensors are connected in parallel at both ends of the battery cells or at the module output to collect current and voltage parameters. Ultrasonic sensors are installed between the cells and inside the module casing to emit and receive high-frequency ultrasonic signals; a wideband impedance tester is connected to the module electrode output to collect impedance characteristics; ambient temperature and humidity sensors are arranged inside the container to monitor ambient temperature and humidity; flow sensors are installed in the coolant lines of the liquid cooling system, and speed sensors are installed at the cooling fan motor of the air cooling system to acquire cooling system operating data; the collected electrical parameters, temperature, environmental, cooling system data, ultrasonic waves, and impedance characteristics are preprocessed and standardized, and then stored in a database, such as... Figure 1 As shown.

[0048] S200, Thermal Stress Modeling: Based on the actual dimensions and component parameters of the battery, a precise 3D solid model of the battery containing each component is created using 3D modeling software. A suitable finite element analysis software and element type are selected, and the element size for different regions of the battery is set to the millimeter level. After determining the discretization strategy, the 3D model is imported for mesh generation, generating a finite element mesh and optimizing its quality. The corresponding thermo-mechanical characteristic parameters of the material are assigned to each micro-element, and the heat exchange and mechanical constraints of the elements are defined. The parameters of the thermal stress solver are set, and the collected temperature data is loaded as a load to complete the preprocessing for thermal stress calculation. A thermal stress model is constructed based on thermoelasticity theory, and the thermal stress of each element is calculated using this model. The calculation formula for the thermal stress model is: Where, σ th Thermal stress represents the stress generated inside the battery due to temperature changes. E is the elastic modulus, the material's ability to resist elastic deformation, obtained from a material database. r is the coefficient of thermal expansion, the material's expansion characteristics with temperature changes, obtained through material testing. ΔT is the temperature difference between the current and previous moments. ν is Poisson's ratio, the ratio of the material's lateral strain to its longitudinal strain, determined by material properties. x, y, and z are spatial coordinates used to locate the positions of the tiny cells inside the battery. The temperature gradient is calculated using the central difference method. The damage threshold is defined by combining this with the critical value of irreversible thermal stress deformation of the battery material. The formula for calculating the damage threshold is: σ cr =σ cr0 ·λ(SOH)·μ(SOC,I), where σ cr It is the critical value for thermal stress damage, the stress threshold at which battery materials undergo irreversible deformation, σ cr0 The initial damage threshold is the baseline damage threshold for a new battery. λ(SOH) is the health state correction coefficient, reflecting the impact of battery aging on the threshold, fitted through cycle life experiments. μ(SOC,I) is the operating condition correction function, obtained through neural network training based on charge / discharge current I and SOC data. SOH represents the battery health state, characterizing the degree of battery capacity decay. SOC is the state of charge, the ratio of remaining battery capacity to rated capacity. I is the charge / discharge current, collected by a current sensor, to determine whether thermal stress will cause irreversible damage to the battery structure. When σ... th <0.8σ cr When the deformation is 0.8σ, it is determined to be elastic deformation, and the structure can recover; cr ≤σ th <σ cr When σ is triggered, a damage warning is issued; when σ is triggered, a damage warning is issued. th ≥σ cr At that time, it was determined to be an irreversible deformation.

[0049] S300, Multi-physics Collaborative Calibration: Simultaneously acquires ultrasonic propagation velocity, AC impedance phase angle, battery structural deformation, and actual contact resistance values. Analyzes the mapping relationships between ultrasonic velocity and structural deformation, and impedance phase angle and contact resistance. Transforms the changes in characteristic quantities into changes in battery structural deformation and contact resistance. Substitute these values ​​into the multi-physics collaborative topology mapping calibration formula to calculate the correction for real-time thermal stress. The calculation formula for the multi-physics collaborative topology mapping calibration is as follows: Where Δσ th This is the thermal stress correction factor, used to correct the deviation of the thermal stress model in real time. ω is the multi-field calibration weight, determined through Bayesian optimization training. The optimization objective is to minimize the thermal stress prediction error. Δv ultra It is the change in ultrasonic velocity, the difference between the current sound velocity and the initial healthy state sound velocity, collected by an ultrasonic sensor, ΔZ. imped It is the impedance change, the difference between the current impedance and the initial healthy state impedance, acquired by a broadband impedance meter. β is the multi-field coupling exponent, which adjusts the weights of the ultrasonic and impedance signals, calibrated through thermal runaway experiments. It is the gradient operator, used to calculate the rate of change of spatial parameters, α. cal It is the calibrated coefficient of thermal expansion, combined with material parameters corrected using multiphysics data, E cal It is the calibrated elastic modulus, combined with material parameters corrected by multiphysics data, ΔS topo It is the entropy change of the topological network, the change in the disorder of the network constructed by multi-physics characteristic quantities, calculated using the information entropy formula, ΔS dam It is the damage entropy change, calculated by the damage entropy increase model in the thermal stress modeling process, where ΔT is the temperature difference between the current moment and the previous moment.

[0050] S400, Dynamic Thermal Model Construction: Based on the traditional heat generation rate model, the influence of thermal stress on the structure and thermal properties is introduced. Combined with calibrated parameters, a dynamic thermal model including a damage diffusion term is constructed. The calculation formula for the dynamic thermal model is as follows: in This is the Caputo fractional enthalpy change rate, describing the historical cumulative effect of enthalpy change. α∈(0,1) is the fractional order, ρ is the cell density, measured by the displacement method, t is the time variable, and c... p It is the constant-pressure heat capacity, obtained by measuring with a calorimeter. T represents the battery temperature. It is a divergence operator used to calculate the spatial distribution of heat flux density, k eff It is the effective thermal conductivity coefficient, including thermal conductivity parameters corrected for structural damage, I 2R is the Joule heat yield, I is the current, R is the battery internal resistance, ξ is the damage diffusion coefficient, calibrated through thermal runaway experiments, characterizing the accelerating effect of damage on heat diffusion, and γ is the damage acceleration index, calibrated in conjunction with η in thermal stress modeling, reflecting the accelerating effect of damage on thermal runaway. It is the Laplace operator, used to calculate the spatial diffusion trend of damage entropy, S dam Damage entropy is calculated using the fractal dimension of the SEM image. ζ is the feedback coefficient, which converts the thermal stress correction into a proportionality coefficient for thermal generation compensation. Δσ th This is the thermal stress correction factor, used to correct deviations in the thermal stress model in real time, σ. cr It is the critical value for thermal stress damage, the stress threshold at which battery materials undergo irreversible deformation. The temperature gradient, where t is a time variable, allows this model to more accurately describe the thermal behavior of a battery during charging and discharging, taking into account the accelerating effect of thermal stress-induced damage on thermal diffusion.

[0051] S500, Thermal Runaway Early Warning Judgment: Construct a multi-dimensional early warning indicator system including temperature, thermal stress, and thermal accumulation trend (reflected by the fractional-order enthalpy change rate output by the dynamic thermal model). Calculate the risk value using a fuzzy-entropy weighted early warning decision function. The calculation formula for the fuzzy-entropy weighted early warning decision is as follows: Where μ i It is the entropy weight coefficient, the indicator weight calculated using information entropy, R. risk This is the thermal runaway risk value; the higher the value, the higher the risk. 'n' represents the number of warning indicators, and 'H' represents the risk level. i H is the information entropy of the i-th indicator. max It is the historical maximum entropy value, Fuzzy(·) is the fuzzy membership function that maps continuous indicators to risk membership degrees, and uses a triangular fuzzy function to define the interval, T. i This is the i-th temperature index data, σ th,i This is the i-th thermal stress index data. It is the fractional-order enthalpy change rate, a thermal accumulation trend index output by the dynamic thermal model. Based on the preset threshold, it triggers corresponding level warnings. When the risk value is less than 0.3, it is low risk, and it is only recorded in the local log of the battery management system and a mild heat dissipation strategy is triggered. When the risk value is between 0.3 and 0.7, it is medium risk, and a warning is pushed to the display terminal and remote monitoring platform, triggering an active temperature equalization strategy and limiting the charging and discharging power to 80%. When the risk value is greater than or equal to 0.7, it is high risk, and an audible and visual alarm, mobile APP push, cloud platform emergency alarm are immediately triggered, and safety intervention measures are linked.

[0052] In summary, this embodiment deploys a multi-source sensor network in a large-scale energy storage power station container scenario, combining thermal stress modeling, multi-field collaborative calibration, and dynamic thermal model construction to form a complete early warning system covering "data acquisition - physical modeling - real-time correction - risk assessment". This scheme utilizes thermoelasticity theory and multi-field collaborative topology mapping calibration formula to fuse and analyze temperature, stress, and impedance. Through fuzzy-entropy weight decision function, it achieves graded early warning of thermal runaway risk, providing a safety protection scheme for high-density energy storage scenarios that balances accuracy and real-time performance, and can effectively prevent the occurrence and spread of thermal runaway accidents.

[0053] Example 2:

[0054] Early warning of thermal runaway of batteries in industrial and commercial energy storage systems.

[0055] An industrial and commercial energy storage system in an industrial park is used for grid peak shaving and power supply within the park. The system consists of multiple battery cabinets, each containing multiple battery modules. Under the complex operating conditions of industrial and commercial power consumption, the batteries frequently charge and discharge, which can easily lead to thermal runaway. This invention can provide effective early warning of battery thermal runaway in the energy storage system, ensuring the safe and reliable power supply to the park.

[0056] S100, Multi-source parameter acquisition: Temperature sensors are installed on the surface of individual battery cells and near the heat dissipation channels of battery modules inside the battery cabinet, such as... Figure 2 As shown, real-time monitoring of battery temperature and heat dissipation involves connecting a current sensor in series in the current loop of the battery module and a voltage sensor in parallel at both ends of the individual battery cells or the module output to collect current and voltage data. Ultrasonic sensors are installed between the cells and inside the module casing to transmit and receive high-frequency ultrasonic signals. A wideband impedance tester is connected to the module electrode output to collect impedance characteristics. Environmental temperature and humidity sensors are placed outside the battery cabinet to monitor ambient temperature and humidity. Flow sensors or speed sensors are installed in the pipes or fan motors of the heat dissipation system to obtain operating parameters. All collected data are preprocessed and standardized before being stored in a database.

[0057] S200, Thermal Stress Modeling: Using 3D modeling software, an accurate 3D solid model of the battery is created according to its actual dimensions and component parameters, including all battery components. Appropriate finite element analysis software and element types are selected. Based on analysis requirements, the element size for different regions of the battery is set to the millimeter level. After determining the discretization strategy, the 3D model is meshed, a finite element mesh is generated, and the mesh quality is checked and optimized. Corresponding material thermo-mechanical property parameters are assigned to each micro-element, and the heat exchange and mechanical constraints of the elements are defined. Relevant parameters of the thermal stress solver are set, and the collected temperature data is used as a load to complete the preprocessing work for thermal stress calculation. A thermal stress model is constructed based on thermoelasticity theory. The calculation formula for the thermal stress model is: The thermal stress of each unit is calculated, and a damage threshold is defined based on the critical value of irreversible deformation of the battery material under thermal stress. The influence of the battery's usage stage (e.g., State of Health (SOH)) and operating conditions (charge / discharge current I, State of Charge (SOC)) on the threshold is also considered. The calculation formula for the thermal stress damage threshold is: σ cr =σ cr0 ·λ(SOH)·μ(SOC,I), to determine whether thermal stress will cause irreversible damage to the battery structure, when σ th <0.8σ cr When the deformation is 0.8σ, it is determined to be elastic deformation, and the structure can recover; cr ≤σ th <σ cr When σ is triggered, a damage warning is issued; when σ is triggered, a damage warning is issued. th ≥σ cr At that time, it was determined to be an irreversible deformation.

[0058] S300, Multi-physics Collaborative Calibration: Simultaneously acquires ultrasonic propagation velocity, AC impedance phase angle, battery structural deformation, and actual contact resistance values. Analyzes the mapping relationships between ultrasonic velocity and structural deformation, and impedance phase angle and contact resistance. Transforms changes in characteristic quantities into changes in battery structural deformation and contact resistance. Utilizes the multi-field collaborative topology mapping calibration formula to calculate the correction for real-time thermal stress. The calculation formula for the multi-field collaborative topology mapping calibration is as follows:

[0059] S400, Dynamic Thermal Model Construction: Based on the traditional heat generation rate model, considering the influence of thermal stress on the battery structure and thermal characteristics, and combining calibrated parameters, a dynamic thermal model including a damage diffusion term is constructed. This model can more comprehensively reflect the heat generation, heat conduction, and damage diffusion processes of the battery during operation. The calculation formula for the dynamic thermal model is as follows: in This is the Caputo fractional enthalpy change rate, describing the historical cumulative effect of enthalpy change. α∈(0,1) is the fractional order, ρ is the cell density, measured by the displacement method, t is the time variable, and c... p It is the constant-pressure heat capacity, obtained by measuring with a calorimeter. T represents the battery temperature. It is a divergence operator used to calculate the spatial distribution of heat flux density, k eff It is the effective thermal conductivity coefficient, including thermal conductivity parameters corrected for structural damage, I 2 R is the Joule heat yield, I is the current, R is the battery internal resistance, ξ is the damage diffusion coefficient, calibrated through thermal runaway experiments, characterizing the accelerating effect of damage on heat diffusion, and γ is the damage acceleration index, calibrated in conjunction with η in thermal stress modeling, reflecting the accelerating effect of damage on thermal runaway. It is the Laplace operator, used to calculate the spatial diffusion trend of damage entropy, S dam Damage entropy is calculated using the fractal dimension of the SEM image. ζ is the feedback coefficient, which converts the thermal stress correction into a proportionality coefficient for thermal generation compensation. Δσ th This is the thermal stress correction factor, used to correct deviations in the thermal stress model in real time, σ. cr It is the critical value for thermal stress damage, the stress threshold at which battery materials undergo irreversible deformation. Temperature gradient, where t is a time variable.

[0060] S500, Thermal Runaway Early Warning Judgment: Establish a multi-dimensional early warning index system including temperature, thermal stress, and thermal accumulation trend (represented by the fractional-order enthalpy change rate output by the dynamic thermal model). Calculate the risk value using a fuzzy-entropy weighted early warning decision function. The calculation formula for the fuzzy-entropy weighted early warning decision is as follows: Where μ i It is the entropy weight coefficient, the indicator weight calculated using information entropy, R. risk This is the thermal runaway risk value; the higher the value, the higher the risk. 'n' represents the number of warning indicators, and 'H' represents the risk level. i H is the information entropy of the i-th indicator. max It is the historical maximum entropy value, Fuzzy(·) is the fuzzy membership function that maps continuous indicators to risk membership degrees, and uses a triangular fuzzy function to define the interval, T. i This is the i-th temperature index data, σ th,i This is the i-th thermal stress index data. It is a fractional-order enthalpy change rate, a thermal accumulation trend index output by the dynamic thermal model. Based on preset thresholds, it triggers different levels of warnings. When the risk value is less than 0.3, it is considered low risk, and is only recorded in the local log of the battery management system, and a mild heat dissipation strategy is initiated. When the risk value is between 0.3 and 0.7, it is considered medium risk, and the warning information is pushed to the display terminal and remote monitoring platform, and an active temperature equalization strategy is initiated, limiting the charging and discharging power to 80%. When the risk value is greater than or equal to 0.7, it is considered high risk, and an audible and visual alarm is immediately issued, an emergency alarm is pushed through the mobile APP and the cloud platform, and relevant safety intervention measures are linked.

[0061] In summary, this embodiment addresses the challenge of thermal runaway early warning under frequent operating condition changes in industrial and commercial energy storage systems through refined parameter acquisition and targeted modeling. Based on thermal stress and dynamic thermal models, and incorporating damage threshold formulas that consider battery health status and operating conditions, as well as multi-field collaborative topology mapping calibration formulas, the solution achieves early identification of internal battery damage. By integrating temperature, stress, and thermal accumulation trend indicators through a fuzzy-entropy weighted early warning decision function, it forms a comprehensive solution from data acquisition to safety intervention. This provides a feasible technical path for the safe operation of industrial and commercial energy storage equipment, demonstrating adaptability and reliability in engineering applications.

[0062] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A method for early warning of thermal runaway damage propagation in energy storage batteries based on a dynamic thermal model, characterized in that, The specific steps of this early warning method are as follows: S100, multi-source parameter acquisition: Multiple sensors are arranged in the battery and module to collect data on electrical parameters, temperature environment and heat dissipation system. Battery material characteristic parameters are obtained from the material database. At the same time, ultrasonic sensors and broadband impedance testers are used to collect multi-physics field characteristic quantities. The collected data are preprocessed, standardized and stored in the database. S200, Thermal Stress Modeling: Based on thermoelasticity theory, the battery is divided into tiny units using the finite element discretization method. A thermal stress model is constructed to calculate the thermal stress of each unit. The damage threshold is defined by combining the critical value of irreversible deformation of thermal stress of battery material. The influence of battery usage stage and operating conditions on the threshold is considered to determine whether thermal stress causes irreversible damage to the battery structure. S300, Multi-physics Field Collaborative Calibration: Simultaneously acquires ultrasonic propagation speed, AC impedance phase angle, battery structure deformation, and actual contact resistance values; analyzes the mapping relationship between ultrasonic speed and structural deformation, and impedance phase angle and contact resistance; transforms the changes in characteristic quantities into changes in battery structure deformation and contact resistance; substitutes these values ​​into the multi-field collaborative topology mapping calibration formula; and corrects the thermal stress in real time. S400, Dynamic Thermal Model Construction: Based on the traditional heat generation rate model, the influence of thermal stress on the structure and thermal properties is introduced, and a dynamic thermal model including damage diffusion term is constructed by combining the calibrated parameters. S500, thermal runaway early warning judgment: Construct a multi-dimensional early warning indicator system that includes temperature, thermal stress, and thermal accumulation trend. Calculate the risk value through a fuzzy-entropy weighted early warning decision function and trigger the corresponding level of early warning according to the preset threshold.

2. The method for early warning of thermal runaway damage propagation in energy storage batteries based on a dynamic thermal model according to claim 1, characterized in that, In S100, the multi-source parameters are collected from sensors arranged in the battery and module, namely: Temperature sensor: installed on the surface of individual battery cells and near the heat dissipation channels of the battery module; Current sensor: installed in series in the current loop of the battery module; Voltage sensor: installed in parallel at both ends of a battery cell or at the output end of a battery module; Ultrasonic sensors: installed between the cells of the battery module and inside the module housing, used to transmit and receive high-frequency ultrasonic signals; Wideband impedance tester: connected to the electrode output terminal of the battery module; Ambient temperature sensor: located on the outside of the battery module; Humidity sensor: installed in the external environment of the battery module; Flow sensor: Installed in the coolant piping of the battery liquid cooling system; Speed ​​sensor: Installed at the cooling fan motor of the battery air-cooling system.

3. The method for early warning of thermal runaway damage propagation in energy storage batteries based on a dynamic thermal model according to claim 1, characterized in that, In the S200 thermal stress modeling process, the specific steps for dividing the battery into micro-units using the finite element discretization method are as follows: (1) Using 3D modeling software, a 3D solid model of the battery containing each component is accurately created based on the actual size and component parameters of the battery. (2) Select finite element analysis software and appropriate element type, and in combination with analysis requirements, set the element size of different areas of the battery to the millimeter level, and determine the discretization strategy. (3) Import the 3D model into the software to perform mesh generation, generate a finite element mesh composed of tiny units, and check and optimize the mesh quality; (4) Assign thermo-mechanical property parameters to each micro-unit, and define the heat exchange and mechanical constraint conditions of the unit; (5) Set the thermal stress solver parameters in the software, load the collected temperature data as the load, and complete the preprocessing for thermal stress calculation.

4. The method for early warning of thermal runaway damage propagation in energy storage batteries based on a dynamic thermal model according to claim 1, characterized in that, S200, the construction of the thermal stress model in thermal stress modeling, the calculation formula of the thermal stress model is: Where, σ th Thermal stress represents the stress generated inside the battery due to temperature changes. E is the elastic modulus, the material's ability to resist elastic deformation, obtained from a material database. r is the coefficient of thermal expansion, the material's expansion characteristics with temperature changes, obtained through material testing. ΔT is the temperature difference between the current and previous moments. ν is Poisson's ratio, the ratio of the material's lateral strain to its longitudinal strain, determined by material properties. x, y, and z are spatial coordinates used to locate the positions of the tiny cells inside the battery. It is the temperature gradient, calculated using the central difference method.

5. The method for early warning of thermal runaway damage propagation in energy storage batteries based on a dynamic thermal model according to claim 1, characterized in that, The calculation formula for the thermal stress damage threshold in S200 thermal stress modeling is: σ cr =σ cr0 ·λ(SOH)·μ(SOC,I), where σ cr It is the critical value for thermal stress damage, the stress threshold at which battery materials undergo irreversible deformation, σ cr0 The initial damage threshold is the baseline damage threshold for a new battery. λ(SOH) is the health state correction coefficient, reflecting the impact of battery aging on the threshold, fitted through cycle life experiments. μ(SOC,I) is the operating condition correction function, obtained through neural network training based on charge / discharge current I and SOC data. SOH represents the battery health state, characterizing the degree of battery capacity decay. SOC is the state of charge, the ratio of remaining battery capacity to rated capacity. I is the charge / discharge current, collected by a current sensor. When σ... th <0.8σ cr When the deformation is 0.8σ, it is determined to be elastic deformation, and the structure can recover; cr ≤σ th <σ cr When σ is triggered, a damage warning is issued; when σ is triggered, a damage warning is issued. th ≥σ cr At that time, it was determined to be an irreversible deformation.

6. The method for early warning of thermal runaway damage propagation in energy storage batteries based on a dynamic thermal model according to claim 5, characterized in that, The calculation formula for the multi-field cooperative topology mapping calibration formula in S300 is as follows: Where Δσ th This is the thermal stress correction factor, used to correct the deviation of the thermal stress model in real time. ω is the multi-field calibration weight, determined through Bayesian optimization training. The optimization objective is to minimize the thermal stress prediction error. Δv ultra It is the change in ultrasonic velocity, the difference between the current sound velocity and the initial healthy state sound velocity, collected by an ultrasonic sensor, ΔZ. imped It is the impedance change, the difference between the current impedance and the initial healthy state impedance, acquired by a broadband impedance meter. β is the multi-field coupling exponent, which adjusts the weights of the ultrasonic and impedance signals, calibrated through thermal runaway experiments. It is the gradient operator, used to calculate the rate of change of spatial parameters, α. cal It is the calibrated coefficient of thermal expansion, combined with material parameters corrected using multiphysics data, E cal It is the calibrated elastic modulus, combined with material parameters corrected by multiphysics data, ΔS topo It is the entropy change of the topological network, the change in the disorder of the network constructed by multi-physics characteristic quantities, calculated using the information entropy formula, ΔS dam It is the damage entropy change, calculated by the damage entropy increase model in the thermal stress modeling process, where ΔT is the temperature difference between the current moment and the previous moment.

7. The method for early warning of thermal runaway damage propagation in energy storage batteries based on a dynamic thermal model according to claim 6, characterized in that, The calculation formula for the dynamic thermal model in the S400 dynamic thermal model construction is as follows: in This is the Caputo fractional enthalpy change rate, describing the historical cumulative effect of enthalpy change. α∈(0,1) is the fractional order, ρ is the cell density, measured by the displacement method, t is the time variable, and c... p It is the constant-pressure heat capacity, obtained by measuring with a calorimeter. T represents the battery temperature. It is a divergence operator used to calculate the spatial distribution of heat flux density, k eff It is the effective thermal conductivity coefficient, including thermal conductivity parameters corrected for structural damage, I 2 R is the Joule heat yield, I is the current, R is the battery internal resistance, ξ is the damage diffusion coefficient, calibrated through thermal runaway experiments, characterizing the accelerating effect of damage on heat diffusion, and γ is the damage acceleration index, calibrated in conjunction with η in thermal stress modeling, reflecting the accelerating effect of damage on thermal runaway. It is the Laplace operator, used to calculate the spatial diffusion trend of damage entropy, S dam Damage entropy is calculated using the fractal dimension of the SEM image. ζ is the feedback coefficient, which converts the thermal stress correction into a proportionality coefficient for thermal generation compensation. Δσ th This is the thermal stress correction factor, used to correct deviations in the thermal stress model in real time, σ. cr It is the critical value for thermal stress damage, the stress threshold at which battery materials undergo irreversible deformation. Temperature gradient, where t is a time variable.

8. The method for early warning of thermal runaway damage propagation in energy storage batteries based on a dynamic thermal model according to claim 7, characterized in that, The calculation formula for the fuzzy-entropy weighted early warning decision in the S500 thermal runaway early warning judgment is as follows: Where μ i It is the entropy weight coefficient, the indicator weight calculated using information entropy, R. risk This is the thermal runaway risk value; the higher the value, the higher the risk. 'n' represents the number of warning indicators, and 'H' represents the risk level. i H is the information entropy of the i-th indicator. max It is the historical maximum entropy value, Fuzzy(·) is the fuzzy membership function that maps continuous indicators to risk membership degrees, and uses a triangular fuzzy function to define the interval, T. i This is the i-th temperature index data, σ th,i This is the i-th thermal stress index data. It is the fractional-order enthalpy change rate, a heat accumulation trend index output by the dynamic thermal model.

9. The method for early warning of thermal runaway damage propagation in energy storage batteries based on a dynamic thermal model according to claim 8, characterized in that, The S500 thermal runaway early warning judgment triggers corresponding level warnings based on preset thresholds: When R risk <0.3 indicates low risk, and will only be recorded in the local logs of the battery management system, triggering a mild heat dissipation strategy; When 0.3≤R risk If the value is less than 0.7, it indicates a medium risk. An early warning will be sent to the display terminal and remote monitoring platform, triggering an active temperature equalization strategy to limit the charging and discharging power to 80%. When R risk A value of ≥0.7 indicates a high risk and will immediately trigger an audible and visual alarm, a push notification to the mobile app, an emergency alarm on the cloud platform, and coordinated safety intervention measures.

Images