Lithium battery thermal runaway internal temperature estimation method based on electrochemical thermal coupling model

Abstract

The invention discloses a lithium battery thermal runaway internal temperature estimation method based on an electrochemical thermal coupling model, and the method comprises the steps: S1, selecting a to-be-measured lithium battery, and obtaining corresponding electrochemical parameters and physical parameters of the to-be-measured lithium battery; s2, carrying out a basic charging and discharging experiment and an extreme working condition experiment on the battery, and collecting related actual measurement data; s3, establishing a battery electrochemical model; s4, establishing a battery thermal model and a battery side reaction model, and obtaining an electrochemical-thermal coupling model; s5, performing numerical simulation on the electrochemical-thermal coupling model to obtain temperature data of each point in the battery at corresponding moments from the early stage to the last stage of thermal runaway of the lithium battery; and S6, using the data in the step S5 and the actually measured data in the step S2 for training a neural network, establishing a neural network model, and predicting the internal temperature of the lithium battery in real time. According to the invention, the advantages of the mechanism model and the data-driven neural network model are combined, and the speed and precision of temperature prediction are improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of power lithium battery applications, and specifically relates to a method for estimating the internal temperature of lithium battery thermal runaway based on an electrochemical-thermal coupling model. Background Art

[0002] At present, China's energy dependence on foreign countries remains high, and the issue of energy security is extremely urgent. To enhance China's energy independence, the country has vigorously developed new energy. However, due to the characteristics of new energy itself, its consumption and supply guarantee have become the primary problems currently faced. In response, energy storage technologies have emerged.

[0003] Energy storage batteries are widely used due to their high energy density and long cycle life. However, when they operate under extreme conditions such as electrical abuse, mechanical pinprick, and thermal abuse, the temperature will rise sharply, leading to thermal runaway. The runaway battery is extremely prone to explosion and releases a huge amount of pressure. The energy storage power station accidents that have occurred in recent years have all caused huge property losses and casualties. Therefore, it is extremely important to monitor the relevant parameters of energy storage batteries and achieve early warning.

[0004] Among a series of parameters, temperature is a relatively critical parameter. Currently, the commonly used methods for monitoring temperature are mainly divided into external temperature monitoring and internal temperature monitoring. However, research has shown that there is a large temperature difference between the inside and outside during thermal runaway, and only monitoring the external temperature will pose risks such as early warning delay. The commonly used internal temperature monitoring methods include directly monitoring through an internal temperature sensor and predicting through an algorithm. However, the internal sensor has disadvantages such as immature technology, easy reduction of battery life, and high cost. Therefore, the present invention proposes a method for estimating the internal temperature of lithium battery thermal runaway based on an electrochemical-thermal coupling model, which improves the prediction speed while reflecting the accuracy. Summary of the Invention

[0005] The purpose of the present invention is to provide a method for estimating the internal temperature of lithium battery thermal runaway based on an electrochemical-thermal coupling model, which is used to solve the problems of inaccurate determination of lithium battery thermal runaway faults and slow early warning speed.

[0006] The present invention is implemented by the following technical solutions:

[0007] A method for estimating the internal temperature of lithium battery thermal runaway based on an electrochemical-thermal coupling model, which specifically includes the following steps:

[0008] Step S1: Select the lithium battery to be tested and obtain its corresponding lithium battery self-characteristic parameters, including electrochemical parameters and physical parameters;

[0009] Step S2: Conduct basic charge and discharge experiments and extreme condition experiments on the lithium battery, and collect relevant measured data;

[0010] Step S3: Establish a battery electrochemical model;

[0011] Step S4: Establish a battery thermal model and a battery side reaction model, and combine them with the battery electrochemical model to finally obtain an electrochemical-thermal coupling model;

[0012] Step S5: Conduct numerical simulation on the electrochemical-thermal coupling model to obtain the temperature data of each point inside the lithium battery at corresponding moments from the early stage to the end stage of thermal runaway;

[0013] Step S6: Use the data in Step S5 and the measured data in Step S2 for the training of a neural network, establish a neural network model, and conduct real-time prediction on the internal temperature of the lithium battery.

[0014] Further preferably, in Step S1, the electrochemical parameters of the lithium battery include the solid-phase lithium-ion concentration in the x-axis direction of the battery, the lithium-ion solid-phase diffusion coefficient of the battery, the liquid-phase volume fraction of the electrolyte, the liquid-phase lithium-ion concentration, the effective liquid-phase lithium-ion diffusion coefficient, the number of lithium-ion jumps, the net interfacial reaction current density, the effective solid-phase conductivity, the solid-phase electric potential, the liquid-phase electric potential, and the exchange current density; the physical parameters of the lithium battery include length, width, height, density, mass, nominal capacity, specific heat capacity, thermal conductivity, and cut-off voltage.

[0015] Further preferably, Step S2 is specifically as follows:

[0016] S21: Conduct basic charge and discharge experiments on the lithium battery to simulate the actual cycling process of the lithium battery;

[0017] S22: Conduct a thermal abuse experiment on the cycled lithium battery to cause it to undergo thermal runaway, and obtain the current, terminal voltage, battery surface temperature, and ambient temperature of the lithium battery under thermal runaway conditions.

[0018] Further preferably, Step S3 is specifically as follows:

[0019] S31: The solid-phase diffusion control equation obtained from the P2D model is:

[0020] ∂ c s ( x , r , t ) ∂ t = D s r 2 · ∂ ∂ r [ r 2 ∂ c s ( x , r , t ) ∂ r ]

[0021] where, is the solid-phase lithium-ion concentration in the x-axis direction of the battery at time t, is the lithium-ion solid-phase diffusion coefficient of the battery, is the particle radial size.

[0022] S32: The liquid-phase diffusion control equation obtained from the P2D model is:

[0023]

[0024] where, is the liquid-phase volume fraction of the electrolyte, is the liquid-phase lithium-ion concentration, is the effective diffusion coefficient of lithium ions in the liquid phase, is the width of the lithium battery, is the number of lithium-ion jumps, is the net current density of the interfacial reaction, is the Faraday constant.

[0025] S33. The solid-phase potential and electron migration process of the solid-phase particles obtained from Ohm's law are as follows:

[0026]

[0027]

[0028] Among them, is the effective solid-phase conductivity, is the solid-phase potential, is the electrode current density, is the specific interfacial area, is the molar flux.

[0029] S34. The liquid-phase potential in the electrolyte is:

[0030]

[0031] Among them, is the effective liquid-phase conductivity, is the liquid-phase potential, is the molar gas constant, is the temperature, is the dimensionless quantity of the electrolyte concentration.

[0032] S35. The electrode reaction kinetic process at the interface of the active particles and the electrolyte is:

[0033] j = i 0 [ exp ( α n F RT η s ) − exp ( α p F RT η s ) ]

[0034] Among them, is the exchange current density, is the negative electrode charge transfer coefficient, is the positive electrode charge transfer coefficient, is the overpotential.

[0035] Further preferably, step S4 is specifically as follows:

[0036] S41: The battery thermal model is a three-dimensional thermal model, including a heat generation model and a heat transfer model; the heat generation model specifically includes a heat balance equation and a total heat generation equation, as follows:

[0037] The heat balance equation is:

[0038]

[0039] wherein, is the density, is the specific heat capacity, is the axial thermal conductivity, is the axial thermal conductivity, is the axial thermal conductivity, is the heat generation; is the width of the lithium battery, is the length of the lithium battery, is the height of the lithium battery;

[0040] The total heat generation equation is:

[0041]

[0042] wherein, is the reversible heat, is the irreversible heat, is the ohmic heat, is the side reaction heat, which is reflected in the following side reaction model;

[0043]

[0044] In the formula, is the open circuit voltage;

[0045] The battery heat transfer model is as follows:

[0046]

[0047] In the formula, is the thermal conductivity, is the normal direction of the battery outer surface, is the heat transfer coefficient, is the battery surface temperature, is the ambient temperature.

[0048] S42. The battery side reaction model is represented by the following formula. When 90°C < T < 120°C, the SEI film decomposes; when T > 120°C, the negative electrode and the electrolyte react; when T > 170°C, the positive electrode and the electrolyte react; when T > 200°C, the electrolyte decomposes;

[0049]

[0050] wherein, is the enthalpy of the decomposition reaction; It is the mass content per unit volume of each active material in the battery; is the reaction rate of each decomposition reaction, where is the reaction rate of SEI film decomposition reaction, is the reaction rate of the negative electrode and electrolyte, is the reaction rate of the positive electrode and electrolyte, is the reaction rate of the decomposition reaction of the electrolyte; is the frequency factor, where is the frequency factor of SEI film decomposition reaction, is the frequency factor of the reaction between the negative electrode and the electrolyte, is the frequency factor of the reaction between the positive electrode and the electrolyte, is the frequency factor of the decomposition reaction of the electrolyte; is the reaction activation energy, where is the activation energy of SEI film decomposition reaction, is the activation energy of the reaction between the negative electrode and the electrolyte, is the activation energy of the reaction between the positive electrode and the electrolyte, is the activation energy of the decomposition reaction of the electrolyte; is the proportion of unstable lithium in the SEI film; is the dimensionless quantity of lithium ion content in the negative electrode carbon layer; is the dimensionless quantity of lithium ion content in the positive electrode carbon layer; is the dimensionless quantity of electrolyte concentration; is the temperature, which is the same as T in the above thermal model.

[0051] S43, combining the battery electrochemical model, the battery thermal model and the battery side reaction model to obtain an electrochemical-thermal coupling model;

[0052] Among them, the battery electrochemical model provides reversible thermal Irreversible heat and Ohm heat ; The battery side reaction model provides side reaction heat for solving the thermal model ; The battery thermal model provides temperature for solving the battery electrochemical model and battery side reaction model The heat calculation results of the battery electrochemical model and the battery side reaction model change the temperature calculation results of the thermal model. Correspondingly, the temperature calculation results in the battery thermal model in turn affect the heat calculation results of the battery electrochemical model and the battery side reaction model. The two influence each other and finally realize model coupling.

[0053] Further preferably, step S5 is specifically as follows: According to the self-characteristic parameters of the lithium battery obtained in step S1 and the current, battery terminal voltage, battery surface temperature, and ambient temperature under the thermal runaway condition of the lithium battery obtained in step S2, substitute them into the electrochemical-thermal coupling model (simultaneous formula), and use the electrochemical-thermal coupling model for numerical simulation. Through the numerical solution calculation tool COMSOL, obtain the temperature data of each point inside the lithium battery at the corresponding moments from the early stage to the end stage of thermal runaway.

[0054] Further preferably, step S6 is specifically as follows:

[0055] S61. Set up a BP neural network model, whose structure includes three layers of neural networks: an input layer, a hidden layer, and an output layer; the inputs of the input layer include the battery terminal voltage U, current I, ambient temperature , battery surface temperature at each moment measured in step S22; there are neurons in the hidden layer; and there are neurons in the output layer, which are respectively the temperature data of each point inside the lithium battery at the corresponding moments from the early stage to the end stage of thermal runaway calculated in step S5.

[0056] S62. Randomly divide the collected data set into three categories: a training set, a test set, and a validation set according to a ratio; the data set includes the battery terminal voltage U, current I, ambient temperature , battery surface temperature at each moment measured in step S22 and the temperature data of each point inside the lithium battery at the corresponding moments from the early stage to the end stage of thermal runaway obtained in step S5.

[0057] S63. Determine the number of neurons in the hidden layer , and its calculation method is as follows:

[0058]

[0059] wherein, is the number of neurons in the input layer, is the number of neurons in the output layer, and is a value between [1, 10].

[0060] S64. Use the function as the activation function of the hidden layer, and its calculation formula is as follows:

[0061]

[0062] wherein, is the input of the function;

[0063] Use the The function is used as the activation function of the output layer and acts as an identity mapping function.

[0064] S65. Collect the battery terminal voltage U, current I, and ambient temperature at each moment , and the battery surface temperature As the input of the BP neural network model, and the temperature at each point inside the lithium battery corresponding to each moment as the output of the BP neural network model for training; the training method includes but is not limited to the backpropagation method, and the training termination condition includes but is not limited to reaching the maximum number of iterations.

[0065] S66. Based on the trained BP neural network model, input the battery terminal voltage U, current I, and ambient temperature under the working conditions to be predicted , and the battery surface temperature , and obtain the temperature prediction values at each point inside the lithium battery corresponding to the working conditions through the BP neural network model.

[0066] The technical solution provided by the present invention has the following advantages compared with the prior art:

[0067] First, mechanism model: Most of the models used in the previous lithium battery internal temperature estimation methods are electrothermal coupling models, which are composed of a first-order equivalent circuit model and a lumped parameter thermal model. However, this model depends on many uncertain internal parameters, such as the positive and negative electrode potentials, conductivity, etc. These parameters often require parameter identification means, and the accuracy of parameter identification directly affects the prediction results of the model, increasing the uncertainty of the model; while the mechanism model adopted in the present invention is an electrochemical-thermal coupling model, which considers the electrochemical reaction process inside the lithium battery, adds side reaction processes that are easily overlooked, and considers the interaction between heat generation and heat transfer. It does not require a complex parameter identification process. The coupled model considers complex internal reactions, improves the accuracy of model calculation, and accurately reflects the temperature distribution inside the lithium battery. Compared with directly measuring the internal temperature of the lithium battery with an internal temperature sensor, it improves the economy.

[0068] Second, neural network model: The BP neural network model can effectively learn the complex non-linear relationships in the data, perform end-to-end learning directly from the original data to the target output, simplifies the model construction process. When dealing with large-scale data, its training process can be accelerated through parallel computing, significantly shortening the training time, avoiding the problem of large computational amount and long iteration time of simply using the mechanism model, shortening the estimation time, and improving the real-time performance of the estimation.

[0069] Thirdly, the present invention estimates the internal temperature during the whole process of lithium battery thermal runaway by using a mechanism model combined with a BP neural network model. By combining the advantages of the mechanism model and the BP neural network model, it not only improves the accuracy of internal temperature estimation but also enhances the estimation speed, helps to identify potential thermal runaway risks, provides an effective method for early warning of thermal runaway, and enhances the safety of lithium batteries.

[0070] The present invention is reasonably designed, combines the advantages of the mechanism model and the data-driven neural network model, improves the speed and accuracy of temperature prediction, and has good practical application value. Brief Description of the Drawings

[0071] Figure 1 It shows the overall flow chart of the method of the present invention. Detailed Embodiments

[0072] In order to more clearly understand the above-mentioned objects, features and advantages of the present invention, the solution of the present invention will be further described below.

[0073] Many specific details are set forth in the following description in order to provide a thorough understanding of the present invention, but the present invention may be practiced in other ways different from those described herein; obviously, the embodiments in the specification are only a part of the embodiments of the present invention, rather than all of the embodiments.

[0074] The specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0075] A method for estimating the internal temperature of lithium battery thermal runaway based on an electrochemical-thermal coupling model provided by an embodiment of the present invention, as Figure 1 shown, includes the following steps:

[0076] Step S1: Select the lithium battery to be tested and obtain its corresponding electrochemical parameters, physical parameters, etc.

[0077] Among them, the electrochemical parameters of the lithium battery include the solid-phase lithium-ion concentration in the x-axis direction of the battery, the solid-phase lithium-ion diffusion coefficient of the battery, the liquid-phase volume fraction of the electrolyte, the liquid-phase lithium-ion concentration, the effective liquid-phase lithium-ion diffusion coefficient, the number of lithium-ion jumps, the net interfacial reaction current density, the effective solid-phase conductivity, the solid-phase potential, the liquid-phase potential, and the exchange current density.

[0078] In the embodiment of the present invention, a ternary lithium battery is used as the simulation object, with a nominal voltage of 3.7V, a battery capacity of 50Ah, a length of 148.3mm, a height of 98mm, a width of 26.7mm, a density of 2311.89 kg / m 3 , a specific heat capacity of 989 J / kg / K, and a thermal conductivity = = 23.3689 W / m / K, = 1.2661 W / m / K, and the cut-off voltage is 2.75 - 4.3 V.

[0079] Step S2: Conduct basic charge-discharge experiments and extreme condition experiments on the lithium battery, and collect relevant measured data.

[0080] S21: Conduct basic charge-discharge experiments on the lithium battery to simulate the actual cycling process of the lithium battery;

[0081] S22: Conduct a thermal abuse experiment on the cycled lithium battery to cause thermal runaway, and obtain the current, battery terminal voltage, battery surface temperature, and ambient temperature of the lithium battery under thermal runaway conditions.

[0082] Step S3: Establish a battery electrochemical model;

[0083] S31: The solid-phase diffusion control equation obtained from the P2D model is:

[0084] ∂ c s ( x , r , t ) ∂ t = D s r 2 · ∂ ∂ r [ r 2 ∂ c s ( x , r , t ) ∂ r ]

[0085] Among them, is the solid-phase lithium-ion concentration in the x-axis direction of the battery at time t, is the lithium-ion solid-phase diffusion coefficient of the battery, is the particle radial size.

[0086] S32: The liquid-phase diffusion control equation obtained from the P2D model is:

[0087]

[0088] Among them, is the liquid-phase volume fraction of the electrolyte, is the liquid-phase lithium-ion concentration, is the effective liquid-phase diffusion coefficient of lithium ions, is the width of the lithium battery, is the number of lithium-ion jumps, is the net interfacial reaction current density, is the Faraday constant.

[0089] S33: The solid-phase potential and electron migration process of the solid-phase particles obtained from Ohm's law are:

[0090]

[0091]

[0092] Among them, is the effective solid-phase conductivity, is the solid-phase electric potential, is the electrode current density, is the specific interfacial area, is the molar flux.

[0093] S34. The liquid-phase electric potential in the electrolyte is:

[0094]

[0095] where, is the effective liquid-phase conductivity, is the liquid-phase electric potential, is the molar gas constant, is the temperature, is the dimensionless quantity of the electrolyte concentration.

[0096] S35. The electrode reaction kinetic process at the interface of the active particles and the electrolyte is:

[0097] j = i 0 [ exp ( α n F RT η s ) − exp ( α p F RT η s ) ]

[0098] where, is the exchange current density, is the negative electrode charge transfer coefficient, is the positive electrode charge transfer coefficient, is the overpotential.

[0099] Step S4. Establish a battery thermal model, a battery side reaction model, and finally establish an electrochemical-thermal coupling model;

[0100] S41: The battery thermal model is a three-dimensional thermal model including a heat generation model and a heat transfer model. The heat generation model specifically includes a heat balance equation and a total heat generation equation, as follows:

[0101] The heat balance equation is:

[0102]

[0103] where, is the density, is the specific heat capacity, is the axial thermal conductivity, is the axial thermal conductivity, is the axial thermal conductivity, is the heat generation amount; is the width of the lithium battery, is the length of the lithium battery, is the height of the lithium battery;

[0104] The total heat generation equation is:

[0105]

[0106] Among them, is the reversible heat, is the irreversible heat, is the ohmic heat, is the side reaction heat, which is reflected in the following side reaction model;

[0107]

[0108] In the formula, is the open circuit voltage;

[0109] The heat transfer model is as follows:

[0110]

[0111] In the formula, is the thermal conductivity, is the normal direction of the outer surface of the battery, is the heat transfer coefficient, is the battery surface temperature, is the ambient temperature.

[0112] S42. The battery side reaction model is represented by the following formula. Among them, when 90°C < T < 120°C, the SEI film decomposes; when T > 120°C, the negative electrode and the electrolyte react; when T > 170°C, the positive electrode and the electrolyte react; when T > 200°C, the electrolyte decomposes;

[0113]

[0114] Among them, is the enthalpy of decomposition reaction; is the mass content per unit volume of each active material in the battery; is the reaction rate of each decomposition reaction, among which is the reaction rate of the SEI film decomposition reaction, is the reaction rate of the negative electrode and the electrolyte reaction, is the reaction rate of the positive electrode and the electrolyte reaction, is the reaction rate of the electrolyte decomposition reaction; is the frequency factor, among which is the frequency factor of the SEI film decomposition reaction, is the frequency factor of the negative electrode and the electrolyte reaction, is the frequency factor of the positive electrode and the electrolyte reaction, is the frequency factor of the electrolyte decomposition reaction; is the reaction activation energy, among which is the activation energy of SEI film decomposition reaction, is the activation energy of the reaction between the negative electrode and the electrolyte, is the activation energy of the reaction between the positive electrode and the electrolyte, is the activation energy of the decomposition reaction of the electrolyte;

[0115] S43, all the formulas of the combined electrochemical model, thermal model and side reaction model are the electrochemical-thermal coupling model, in which the battery electrochemical model provides reversible thermal energy for solving the thermal model. Irreversible heat and Ohm heat ; The battery side reaction model provides side reaction heat for solving the thermal model ; The battery thermal model provides temperature for solving the electrochemical model and side reaction model The heat calculation results of the battery electrochemical model and the battery side reaction model change the temperature calculation results of the battery thermal model. Correspondingly, the temperature calculation results in the battery thermal model in turn affect the heat calculation results of the battery electrochemical model and the battery side reaction model. The two influence each other and finally realize model coupling.

[0116] Step S5, numerically simulating the electrochemical-thermal coupling model to obtain the temperature data of each point inside the lithium battery at the corresponding time from the early stage to the late stage of thermal runaway;

[0117] According to the characteristic parameters of the lithium battery itself obtained in step S1 and the current, battery terminal voltage, battery surface temperature, and ambient temperature of the lithium battery under thermal runaway conditions obtained in step S22, the above-mentioned combined formulas are substituted, and numerical simulation is performed using the electrochemical-thermal coupling model. The temperature data of each point inside the lithium battery at the corresponding moments from the early stage to the late stage of thermal runaway are obtained through the numerical solution calculation tool COMSOL.

[0118] Step S6: Use the data from step S5 and the measured data from step S2 to train a neural network, establish a neural network model, and perform real-time prediction of the internal temperature of the lithium battery. Specifically, the following steps are included:

[0119] S61, set up a BP neural network model, whose structure includes a three-layer neural network of input layer, hidden layer and output layer; the input of the input layer includes the battery terminal voltage U, current I, ambient temperature at each moment , Battery surface temperature Four neurons; the hidden layer has neurons; the output layer has The neurons are the temperature data of each point inside the lithium battery at the corresponding time from the early stage to the late stage of thermal runaway obtained by calculation in step S5.

[0120] S62. Randomly divide the collected data set according to the ratio of 75% for the training set, 15% for the test set, and 15% for the validation set. Among them, the data set includes the battery terminal voltage U, current I, and ambient temperature at each moment measured in step S22 , the battery surface temperature and the temperature data of each point inside the lithium battery at the corresponding moments from the early stage to the end stage of thermal runaway obtained in step S5.

[0121] S63: Determine the number of neurons in the hidden layer , and its calculation method is as follows:

[0122]

[0123] Among them, is the number of neurons in the input layer, is the number of neurons in the output layer, is a value between [1, 10]. In the embodiment of the present invention, = 4, = 4, and the number of neurons in the hidden layer is taken as 10.

[0124] S64. Use the function as the activation function of the hidden layer, and its calculation formula is as follows:

[0125]

[0126] Among them, is the input of the function;

[0127] Use the function as the activation function of the output layer, which is used as the identity mapping function.

[0128] S65. Collect the battery terminal voltage U, current I, and ambient temperature at each moment , the battery surface temperature as the input of the BP neural network model, and the temperature of each point inside the lithium battery at the corresponding moment as the output of the BP neural network model for training. The training method in the embodiment of the present invention uses the L-M algorithm (Levenberg-Marquardt). When the maximum number of iterations is 1500 or the mean square error MSE is lower than 1e -5 , the training is terminated.

[0129] S66. Based on the trained BP neural network model, input the battery terminal voltage U, current I, and ambient temperature , the battery surface temperature under the working condition to be predicted, and obtain the temperature prediction value of each point inside the lithium battery under the corresponding working condition through the BP neural network model.

[0130] The above are only specific embodiments of the present invention, enabling those skilled in the art to understand or implement the present invention. Although the foregoing embodiments have been described in detail, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some or all of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments, and they should all be covered by the protection scope of the claims.

Claims

1. A method for estimating the internal temperature of a lithium battery thermal runaway based on an electrochemical thermal coupling model, characterized in that: It includes the following steps: Step S1: Select the lithium battery to be tested and obtain the corresponding self-characteristics parameters of the lithium battery, including electrochemical parameters and physical parameters; Step S2: Conduct basic charge and discharge experiments and extreme condition experiments on the lithium battery, and collect relevant measured data; Step S3: Establish a battery electrochemical model; Step S4: Establish a battery thermal model and a battery side reaction model, and combine them with the battery electrochemical model to finally obtain an electrochemical-thermal coupling model; Step S5: Conduct numerical simulation on the electrochemical-thermal coupling model to obtain the temperature data of each point inside the lithium battery at corresponding moments from the early stage to the end stage of thermal runaway; Step S6: Use the data in Step S5 and the measured data in Step S2 for the training of a neural network, establish a neural network model, and conduct real-time prediction on the internal temperature of the lithium battery.

2. The method for estimating the internal temperature of a lithium battery under thermal runaway based on an electrochemical thermal coupling model according to claim 1, characterized in that: In Step S1, the electrochemical parameters of the lithium battery include the solid-phase lithium-ion concentration in the x-axis direction of the battery, the solid-phase lithium-ion diffusion coefficient of the battery, the liquid-phase volume fraction of the electrolyte, the liquid-phase lithium-ion concentration, the effective liquid-phase lithium-ion diffusion coefficient, the number of lithium-ion jumps, the net interfacial reaction current density, the effective solid-phase conductivity, the solid-phase electric potential, the liquid-phase electric potential, and the exchange current density; the physical parameters of the lithium battery include length, width, height, density, mass, nominal capacity, specific heat capacity, thermal conductivity, and cut-off voltage.

3. The method for estimating the internal temperature of a lithium battery under thermal runaway based on an electrochemical thermal coupling model according to claim 2, characterized in that: Step S2 is as follows: S21: Conduct basic charge and discharge experiments on the lithium battery to simulate the actual cyclic use process of the lithium battery; S22: Conduct a thermal abuse experiment on the cycled lithium battery to cause it to undergo thermal runaway, and obtain the current, battery terminal voltage, battery surface temperature, and ambient temperature of the lithium battery under thermal runaway conditions.

4. The method for estimating the internal temperature of a lithium battery under thermal runaway based on an electrochemical thermal coupling model according to claim 3, characterized in that: Step S3 is as follows: S31: The solid-phase diffusion control equation obtained from the P2D model is: ； in, is the solid phase lithium ion concentration in the x-axis direction of the battery at time t, is the solid phase diffusion coefficient of lithium ions in the battery, is the radial size of the particle; S32: The liquid-phase diffusion control equation obtained from the P2D model is: ； in, is the electrolyte liquid phase volume fraction, is the liquid phase lithium ion concentration, is the effective diffusion coefficient of lithium ions in liquid phase, is the width of the lithium battery, is the lithium ion transition number, is the net current density of the interfacial reaction, is the Faraday constant; S33: The solid-phase electric potential and electron migration process of the solid-phase particles obtained from Ohm's law are: ； ； in, is the effective solid phase conductivity, is the solid phase potential, is the electrode current density, is the specific surface area, is the molar flux; S34: The liquid-phase electric potential in the electrolyte is: ； in, is the effective liquid conductivity, is the liquid phase potential, is the molar gas constant, is the temperature, is the dimensionless quantity of electrolyte concentration; S35: The electrode reaction kinetic process at the interface between the active particles and the electrolyte is: ； in, is the exchange current density, is the negative electrode charge transfer coefficient, is the positive electrode charge transfer coefficient, is the overpotential.

5. The method for estimating the internal temperature of a lithium battery under thermal runaway based on an electrochemical thermal coupling model according to claim 4, characterized in that: Step S4 is as follows: S41: The battery thermal model is a three-dimensional thermal model, including a heat generation model and a heat transfer model; The heat generation model includes a heat balance equation and a total heat generation equation, which are specifically as follows: The heat balance equation is: ； in, is the density, is the specific heat capacity, for Axial thermal conductivity, for Axial thermal conductivity, for Axial thermal conductivity, To produce heat; is the width of the lithium battery, is the length of lithium battery, is the height of lithium battery; The total heat generation equation is: ； in, is reversible heat, is irreversible heat, For Ohmic heat, Heat of side reaction; ； In the formula, is the open circuit voltage; The heat transfer model is as follows: ； In the formula, is the thermal conductivity, is the normal direction of the battery outer surface, is the heat transfer coefficient, is the battery surface temperature, is the ambient temperature; S42: The battery side reaction model is represented by the following formula, where when 90°C < T < 120°C, the SEI film decomposes; when T > 120°C, the negative electrode reacts with the electrolyte; when T > 170°C, the positive electrode reacts with the electrolyte; when T > 200°C, the electrolyte decomposes; ； in, is the enthalpy of decomposition reaction; It is the mass content per unit volume of each active material in the battery; is the reaction rate of each decomposition reaction, where is the reaction rate of SEI film decomposition reaction, is the reaction rate of the negative electrode and electrolyte, is the reaction rate of the positive electrode and electrolyte, is the reaction rate of the decomposition reaction of the electrolyte; is the frequency factor of SEI film decomposition reaction, is the frequency factor of the reaction between the negative electrode and the electrolyte, is the frequency factor of the reaction between the positive electrode and the electrolyte, is the frequency factor of the decomposition reaction of the electrolyte; is the activation energy of SEI film decomposition reaction, is the activation energy of the reaction between the negative electrode and the electrolyte, is the activation energy of the reaction between the positive electrode and the electrolyte, is the activation energy of the decomposition reaction of the electrolyte; is the proportion of unstable lithium in the SEI film; is the dimensionless quantity of lithium ion content in the negative electrode carbon layer; is the dimensionless quantity of lithium ion content in the positive electrode carbon layer; is the dimensionless quantity of electrolyte concentration; is temperature; S43: After联立 the battery electrochemical model, the battery thermal model, and the battery side reaction model, an electrochemical-thermal coupling model is obtained; Among them, the battery electrochemical model provides reversible thermal energy for solving the battery thermal model. Irreversible heat and Ohm heat ; The battery side reaction model provides side reaction heat for solving the battery thermal model ; The battery thermal model provides temperature for solving the battery electrochemical model and battery side reaction model .

6. The method for estimating the internal temperature of a lithium battery under thermal runaway based on an electrochemical thermal coupling model according to claim 5, characterized in that: Step S5 is as follows: According to the self-characteristics parameters of the lithium battery obtained in Step S1 and the current, battery terminal voltage, battery surface temperature, and ambient temperature of the lithium battery under thermal runaway conditions obtained in Step S2, substitute them into the electrochemical-thermal coupling model, and use the electrochemical-thermal coupling model for numerical simulation to obtain the temperature data of each point inside the lithium battery at corresponding moments from the early stage to the end stage of thermal runaway.

7. The method for estimating the internal temperature of a lithium battery under thermal runaway based on an electrochemical thermal coupling model according to claim 6, characterized in that: Step S6 is as follows: S61, set up a BP neural network model, whose structure includes a three-layer neural network of input layer, hidden layer and output layer; the input of the input layer includes the battery terminal voltage U, current I, ambient temperature at each moment measured in step S22 , Battery surface temperature four neurons; the hidden layer has neurons; the output layer has neurons, which are the temperature data of each point inside the lithium battery at the corresponding time from the early stage to the end stage of thermal runaway obtained by calculation in step S5; S62, randomly divide the collected data set into three categories: training set, test set, and validation set according to proportion; the data set includes the battery terminal voltage U, current I, and ambient temperature at each moment measured in step S22 , Battery surface temperature and the temperature data of each point inside the lithium battery at the corresponding time from the early stage to the late stage of thermal runaway obtained in step S5; S63. Determine the number of neurons in the hidden layer , which is calculated as follows: ； in, is the number of neurons in the input layer, is the number of neurons in the output layer, is a value between [1,10]; S64, will The function is used as the hidden layer activation function, and its calculation formula is as follows: ； in, for Function input; Will The function is used as the output layer activation function as the identity mapping function; S65: Collect the battery terminal voltage U, current I, and ambient temperature at each moment , Battery surface temperature As the input of the BP neural network model, the temperature of each point inside the lithium battery at each moment is used as the output of the BP neural network model for training; S66, based on the trained BP neural network model, input the battery terminal voltage U, current I, and ambient temperature under the working condition to be predicted , Battery surface temperature , the temperature prediction value of each point inside the lithium battery under the corresponding working conditions is obtained through the BP neural network model.

8. The method for estimating the internal temperature of a lithium battery under thermal runaway based on an electrochemical thermal coupling model according to claim 7, characterized in that: In step S65, the training method is the back propagation method, and the training termination condition is reaching the maximum number of iterations.

Images