High-capacity energy storage lithium ion battery module surface temperature field prediction method and system

Abstract

The invention discloses a high-capacity energy storage lithium ion battery module surface temperature field prediction method and system. The method comprises the following steps: acquiring temperature values of temperature measuring points at all pressure release valves of a to-be-predicted lithium ion battery module; and inputting the obtained temperature value into a preset physical information-neural network model for prediction to obtain a temperature field of the surface of the lithium ion battery module to be predicted. The system comprises a data acquisition module and a data processing module. The invention provides reliable technical support for thermal management of the battery module, and has important engineering application value.

Description

Technical Field

[0001] The present invention belongs to the technical field of lithium - ion batteries, and particularly relates to a method and system for predicting the surface temperature field of a large - capacity energy - storage lithium - ion battery module. Background Art

[0002] Lithium - ion batteries are widely used in energy - storage systems due to their high energy density and power density. However, temperature fluctuations, especially uneven temperature distribution, can significantly affect the performance, safety, and lifespan of battery modules. Uneven temperature can lead to capacity fade, thermal runaway, and non - uniform aging of individual cells within the battery module, thus accelerating the aging process and limiting system performance. With the increase in battery capacity and size, accurately predicting the temperature field of large - capacity battery modules has become a key challenge.

[0003] Existing research mainly focuses on temperature prediction and thermal management of individual cells. For example, Xie et al. proposed an adaptive 3D thermal model based on the resistance transfer algorithm (RTA) and thermal resistance network, significantly improving the calculation efficiency. Dai et al. achieved high - precision estimation of internal battery temperature and external thermal resistance based on the Kalman filter and an equivalent time - varying electrical network thermal model. Richardson et al. proposed a radial 1D model based on electrochemical impedance and surface temperature to estimate the internal temperature distribution without the need for battery thermal characteristics or thermal boundary conditions. In addition, Xu et al. and Naguib et al. used machine - learning methods to predict the surface temperature of individual cells, reducing the dependence on temperature sensors, but errors still exist.

[0004] For temperature prediction of battery modules, existing research mostly focuses on small - capacity batteries or electric - vehicle battery packs. Ranjan et al. used ANN, RNN, and LSTM to estimate the temperature of electric - vehicle battery packs and found that the ANN model had the highest accuracy. Kim et al. proposed a sensorless temperature prediction method based on short - time - series data of voltage and discharge current, ensuring high accuracy and fast execution through an adaptive sequence - length strategy. Cho et al. combined a physical model and a data - driven model to propose a physics - informed neural network method, accurately predicting the temperature of a 70Ah lithium - ion battery module.

[0005] However, existing technologies still have deficiencies in the temperature prediction of large - capacity battery modules. Traditional temperature sensors cannot comprehensively capture the overall thermal distribution and temperature gradient of battery modules, especially in the inter - battery transition region of large - capacity batteries. In addition, existing methods mostly rely on a large number of sensors, increasing system complexity and cost. Although the PINN method performs well in solving complex mathematical equations and lithium - ion battery models, its application in predicting the temperature field of large - capacity battery modules still requires further research. Therefore, developing a PINN - based method for predicting the temperature field of large - capacity battery modules is of great significance for optimizing thermal management strategies and improving system safety and lifespan. Summary of the Invention

[0006] The present invention provides a method and system for predicting the surface temperature field of a large-capacity energy storage lithium-ion battery module, aiming to achieve temperature prediction for large-capacity battery modules.

[0007] To achieve the above object, the present invention adopts the following technical solutions:

[0008] A method for predicting the surface temperature field of a large-capacity energy storage lithium-ion battery module, comprising:

[0009] Obtaining the temperature values of the temperature measurement points at each pressure relief valve of the lithium-ion battery module to be predicted;

[0010] Inputting the obtained temperature values into a preset physics-informed neural network model for prediction to obtain the temperature field on the surface of the lithium-ion battery module to be predicted; wherein, the preset physics-informed neural network model includes: simplification and modeling of the heat transfer physics problem of the battery module, the basic neural network part and the physics information fusion part. The physics information fusion part adds the battery heat transfer differential equation and the battery boundary conditions as physical information constraints during the neural network training process. The loss function in the neural network includes boundary loss, partial differential equation loss, and observation dataset loss.

[0011] A further improvement of the present invention lies in that the simplification and modeling of the heat transfer physics problem of the battery module includes:

[0012] For the convenience of modeling, the battery module is conceptually unfolded with the top surface as the central reference plane. Through this transformation, the temperature distributions of 11 surfaces represent the surface temperature of the battery module; the steady-state temperature distribution of the battery is controlled by the following equation:

[0013]

[0014] When solving the heat transfer problem of the battery module, the heat conduction mechanism inside a single battery and between adjacent batteries is analyzed preferentially; through the layout of heat sources and the focused analysis of the heat conduction mechanism, the calculation of the battery temperature field is simplified, and the best balance is achieved between the acceleration of the training process and the calculation accuracy.

[0015] A further improvement of the present invention lies in that the temperature field prediction based on the physics-informed neural network includes:

[0016] In the physics-informed neural network model, first estimate the initial heating power according to the approximate temperature distribution of the battery, and then train the weights and biases of the model based on the heat balance equation; for each heating unit, combine the heat conduction physical processes inside and between the batteries to construct the loss function defined by equations (2)-(4):

[0017]

[0018] Loss bound = Loss bound_x + Loss bound_y + Loss bound_z (3)

[0019]

[0020] wherein, W obs , and W bound respectively represent the weight parameters of the observed data, partial differential equation, and boundary conditions; Loss pde , Loss obs , Loss bound , Loss respectively represent the partial differential equation loss, observed data loss, boundary condition loss, and total loss; Loss bound_x , Loss bound_y , Loss bound_z respectively represent the boundary losses in the x, y, and z directions; wherein, the thermal conductivities in the x, y, and z directions are set to 10.63 W / (m·K), 2.55 W / (m·K), and 7.74 W / (m·K) respectively, and the heat transfer coefficient between the battery surface and the air is 10 W / (m 2 ·K); by minimizing this loss function, the MLP network is trained to predict the battery temperature.

[0021] A further improvement of the present invention lies in that in the preset physical information - neural network model, the basic neural network part is constructed of a multi - layer perceptron; the physical information fusion part refers to the part where the loss function in the neural network is minimized during training.

[0022] A further improvement of the present invention lies in that during the construction of the multi - layer perceptron, the open - source deep learning framework PyTorch is used; in the physical information - neural network model, the depth of the multi - layer perceptron MLP is 6, which is composed of an input layer, four hidden layers, and an output layer respectively, and each layer contains 100 neurons; the tanh function is selected as the activation function; the initial learning rate is set to 1e - 4, and the weights of the partial differential equation, boundary conditions, and observed data are set to 1, 1, and 1e - 4 respectively.

[0023] A further improvement of the present invention lies in that the input of the physical information - neural network model is the temperature data measured by the experimentally verified battery limited pressure relief valve. At this time, the loss function is the sum of the boundary loss, partial differential equation loss, and observed data set loss. Among them, the boundary condition loss term includes the heat transfer within the battery module and between the module and the air. In the partial differential equation loss, the approximate range of the temperature of some parts of the battery based on experimental data is added. By minimizing Loss, the network parameters are trained, and finally the optimal parameters of the battery temperature reconstruction model are obtained, and then the predicted temperature field parameter set is obtained.

[0024] Prediction system for temperature field of large-capacity energy storage lithium-ion battery module, comprising:

[0025] Data acquisition module, which acquires the temperature values of temperature measurement points at each pressure relief valve of the lithium-ion battery module to be predicted;

[0026] Data processing module, which inputs the acquired temperature values into a preset physical information-neural network model for prediction to obtain the temperature field on the surface of the lithium-ion battery module to be predicted; wherein, the preset physical information-neural network model includes: simplification and modeling of heat transfer physical problems of the battery module, basic neural network part and physical information fusion part. In the physical information fusion part, the heat transfer differential equation of the battery and the battery boundary conditions are added as physical information constraints during the neural network training process, and the loss function in the neural network includes boundary loss, partial differential equation loss and observation data set loss.

[0027] A further improvement of the present invention lies in that in the data acquisition module, acquiring the temperature values of a small number of temperature measurement points of the lithium-ion battery module to be predicted includes:

[0028] Based on the battery temperature measurement experimental platform, temperature measurement experiments are carried out on a three-series 280Ah battery module under different working conditions to obtain the temperature measurement values at the pressure relief valves on the surface of the battery module, which are used as the temperature values of a small number of temperature measurement points of the lithium-ion battery module to be predicted.

[0029] A further improvement of the present invention lies in that in the data processing module, the simplification and modeling of heat transfer physical problems of the battery module include:

[0030] For the convenience of modeling, the battery module is conceptually unfolded with the top surface as the central reference plane. Through this conversion, the temperature distribution on 11 surfaces represents the surface temperature of the battery module; the steady-state temperature distribution of the battery is controlled by the following equation:

[0031]

[0032] When solving the heat transfer problem of the battery module, the heat conduction mechanism inside a single battery and between adjacent batteries is analyzed preferentially; through the layout of heat sources and the focused analysis of the heat conduction mechanism, the calculation of the battery temperature field is simplified, and the best balance is achieved between the acceleration of the training process and the calculation accuracy.

[0033] A computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, it realizes the steps of a prediction method for the temperature field of a large-capacity energy storage lithium-ion battery module as described above.

[0034] Compared with the prior art, the present invention has at least the following beneficial technical effects:

[0035] The present invention adopts a physical information-neural network model, providing an efficient and accurate solution for the temperature field prediction of a three-series 280Ah battery module. By directly embedding the heat transfer physical laws of the battery module into the neural network training process, this method realizes the organic combination of physical constraints and data-driven. In the modeling process, through the reasonable simplification of the heat source distribution of the battery module and the key analysis of the heat conduction mechanism, the computational complexity is significantly reduced while ensuring the physical consistency of the model. In the training stage, by optimizing the composite loss function including physical equations, boundary conditions, and observation data, the network can accurately learn the temperature distribution law of the battery module. Compared with traditional methods, the advantages of the present invention are as follows: First, by introducing physical information, the dependence on a large amount of training data is reduced, and high-precision prediction can be achieved by using only individual temperature measurement points; Second, while ensuring the prediction accuracy, the computational efficiency is greatly improved, and it can quickly respond to the temperature change of the battery module; Third, the introduction of physical constraints enhances the generalization ability of the model, enabling it to maintain stable prediction performance under different working conditions. This method provides reliable technical support for the thermal management of battery modules and has important engineering application value. Description of the Drawings

[0036] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required for use in the description of the specific embodiments or the prior art. Obviously, the following drawings are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0037] Figure 1 It is the overall flowchart in an embodiment of the present invention;

[0038] Figure 2 It is the developed view of the battery module modeling in an embodiment of the present invention;

[0039] Figure 3 It is the schematic diagram of the surface temperature field of the battery module predicted by the PINN algorithm at 1C in an embodiment of the present invention;

[0040] Figure 4 It is the structural block diagram of the prediction system for the temperature field of large-capacity energy storage lithium-ion batteries in an embodiment of the present invention. Detailed Embodiments

[0041] In the following text, only some exemplary embodiments are simply described. As those skilled in the art can recognize, the described embodiments can be modified in various different ways without departing from the spirit or scope of the present invention. Therefore, the drawings and the description are considered to be exemplary in nature rather than restrictive.

[0042] In the description of the present invention, it should be understood that when used in this specification and the appended claims, the terms "comprising" and "including" indicate the presence of the described features, wholes, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components, and / or their combinations.

[0043] It should also be understood that the terms used in the specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. As used in the specification of the present invention and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an", and "the" are intended to include the plural forms.

[0044] It should be further understood that the term " / and" used in the specification of the present invention and the appended claims refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0045] Schematic diagrams of various structures according to the disclosed embodiments of the present invention are shown in the drawings. These figures are not drawn to scale, where for the purpose of clear expression, some details are enlarged and some details may be omitted. The shapes of various regions and layers shown in the figures and their relative sizes and positional relationships are only exemplary, and may actually deviate due to manufacturing tolerances or technical limitations, and those skilled in the art can design regions / layers with different shapes, sizes, and relative positions according to actual needs.

[0046] The embodiments of the present invention will be described in detail below with reference to the drawings.

[0047] Embodiment 1

[0048] To achieve high-precision prediction of the temperature field of large-capacity battery cells, the present invention conducts in-depth analysis based on the heat transfer characteristics of the battery, combines the physics-informed neural network model, and proposes a high-precision prediction method for the surface of the battery module, as Figure 1 shown. In this method, the two-dimensional coordinates of the battery surface are used as the input parameters of the PINN model. By optimizing the combined loss function, the training of the PINN model is realized, so as to ensure that it can accurately predict the temperature field distribution of the battery. In particular, the key physical information of the internal heat transfer of the battery module is incorporated into the control equation and boundary conditions, further improving the prediction accuracy of the model. This method realizes the rapid and accurate prediction of the temperature distribution of the three-series 280Ah battery module by effectively using the temperature measurement data of three pressure relief valves, providing reliable technical support for battery thermal management. The following will introduce each key link of the present invention one by one in the form of embodiments:

[0049] (1) Simplification and Modeling of Heat Transfer Physical Problems in Battery Modules

[0050] For ease of modeling, the battery module (excluding the bottom surface) is conceptually unfolded with the top surface as the central reference plane, as Figure 2 shown. Through this transformation, Figure 2 the temperature distribution of the 11 surfaces in

[0051]

[0052] can effectively represent the surface temperature of the battery module (excluding the bottom surface). Therefore, the steady-state temperature distribution T of the battery is controlled by the following equation:

[0053] (2) Temperature Field Prediction Based on Physics-Informed Neural Network

[0054] In the proposed PINN model, the initial heating power is first estimated based on the approximate temperature distribution of the battery, and then the weights and biases of the model are trained based on the heat balance equation. For each heating unit, a loss function defined by equations (2)-(4) is constructed by combining the physical processes of heat conduction inside and between the batteries.

[0055]

[0056] Loss bound = Loss bound_x + Loss bound_y + Loss bound_z (3)

[0057]

[0058] where W obs , and W bound respectively represent the weight parameters of the observed data, partial differential equations, and boundary conditions; Loss pde , Loss obs , Loss bound , Loss respectively represent the partial differential equation loss, observed data loss, boundary condition loss, and total loss; Loss bound_x , Loss bound_y , Loss bound_zrespectively represent the boundary losses in the x, y, and z directions; among them, the thermal conductivities in the x, y, and z directions are respectively set to 10.63 W / (m·K), 2.55 W / (m·K), and 7.74 W / (m·K), and the heat transfer coefficient between the battery surface and the air is 10 W / (m 2 ·K); by minimizing this loss function, the MLP network is trained to predict the battery temperature T α .

[0059] Example 2

[0060] Experimental verification for 280 Ah battery

[0061] In the present invention, a lithium iron phosphate battery module with a capacity of 280 Ah in three series is selected as the research object. In order to verify the accuracy of the PINN temperature reconstruction algorithm, the temperature observation points are set at the pressure relief valves of three battery cells. As Figure 3 shown is the battery surface temperature field predicted by the PINN algorithm when the battery is charged at a 1C condition for 3600 s.

[0062] Through the above steps, a method for predicting the surface temperature field of a 280 Ah large-capacity energy storage lithium-ion battery module based on physics-informed neural network can be established.

[0063] Example 3

[0064] As Figure 4 shown, the prediction system for the temperature field of the large-capacity energy storage lithium-ion battery module provided by the present invention includes:

[0065] A data acquisition module that acquires the temperature values of the temperature measurement points at the pressure relief valves of the lithium-ion battery module to be predicted;

[0066] A data processing module that inputs the acquired temperature values into a preset physics-informed neural network model for prediction to obtain the temperature field on the surface of the lithium-ion battery module to be predicted; among them, the preset physics-informed neural network model includes: simplification and modeling of the heat transfer physical problem of the battery module, the basic neural network part and the physical information fusion part. The physical information fusion part adds the battery heat transfer differential equation and the battery boundary conditions as physical information constraints during the neural network training process. The loss function in the neural network includes boundary loss, partial differential equation loss, and observation dataset loss.

[0067] In this embodiment, in the data acquisition module, acquiring the temperature values of a small number of temperature measurement points of the lithium-ion battery module to be predicted includes:

[0068] Based on the battery temperature measurement experimental platform, temperature measurement experiments are carried out on a three-series 280Ah battery module under different working conditions, and the temperature measurement values at the pressure relief valve on the surface of the battery module are obtained as the temperature values of a small number of temperature measurement points of the lithium-ion battery module to be predicted.

[0069] In this embodiment, in the data processing module, the simplification and modeling of the heat transfer physical problems of the battery module include:

[0070] For the convenience of modeling, the battery module is conceptually unfolded with the top surface as the central reference plane. Through this transformation, the temperature distribution of 11 surfaces represents the surface temperature of the battery module; the steady-state temperature distribution of the battery is controlled by the following equation:

[0071]

[0072] When solving the heat transfer problem of the battery module, the heat conduction mechanism inside a single battery and between adjacent batteries is analyzed preferentially; through the layout of heat sources and the focused analysis of the heat conduction mechanism, the calculation of the battery temperature field is simplified, and the best balance is achieved between the acceleration training process and the calculation accuracy.

[0073] Embodiment 4

[0074] A computer-readable storage medium provided by the present invention stores a computer program, and when the computer program is executed by a processor, the steps of the method for predicting the temperature field of a large-capacity energy storage lithium-ion battery module are implemented.

[0075] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0076] The present application is described with reference to the flowcharts and / or block diagrams of methods, systems, and computer program products according to the embodiments of the present application. It should be understood that each process and / or block in the flowcharts and / or block diagrams, and the combination of processes and / or blocks in the flowcharts and / or block diagrams can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate for implementing in the process Figure 1 each process or multiple processes and / or blocks Figure 1A system with functions specified in one or more boxes.

[0077] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, such that the instructions stored in the computer-readable memory produce a manufactured article including an instruction device that implements the functions specified in one Figure 1 one process or more processes and / or boxes Figure 1 one box or more boxes.

[0078] These computer program instructions can also be loaded onto a computer or other programmable data processing device, such that a series of operation steps are executed on the computer or other programmable device to produce a computer-implemented process. Thus, the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one Figure 1 one process or more processes and / or boxes Figure 1 one box or more boxes.

[0079] The inventive points protected by the present invention:

[0080] 1. Using physical information-neural network, a method for predicting the surface temperature field of a 280Ah large-capacity energy storage lithium-ion battery module based on physical information-neural network is proposed. High-precision and rapid prediction of the temperature field can be achieved only with the measurement points at the pressure relief valve.

[0081] 2. Experimental verification is carried out using a 280Ah three-series battery module under 1C working conditions.

[0082] The above has shown and described the basic principles, main features and advantages of the present invention. For those skilled in the art, it is obvious that the present invention is not limited to the details of the above exemplary embodiments, and without departing from the spirit or basic features of the present invention, the present invention can be implemented in other specific forms. Therefore, from any point of view, the embodiments should be regarded as exemplary and non-limiting. The scope of the present invention is defined by the appended claims rather than the above description. Therefore, all changes falling within the meaning and scope of the equivalent elements of the claims are intended to be included in the present invention. Any reference signs in the claims should not be regarded as limiting the claimed rights.

[0083] In addition, it should be understood that although this specification is described in terms of embodiments, not every embodiment contains only an independent technical solution. This narrative manner of the specification is only for clarity. Those skilled in the art should regard the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art. The above content is only to illustrate the technical idea of the present invention and cannot be used to limit the protection scope of the present invention. Any modification made on the basis of the technical solution according to the technical idea proposed by the present invention falls within the protection scope of the claims of the present invention.

Claims

1. A method for predicting the surface temperature field of a large-capacity energy storage lithium-ion battery module, characterized in that, Including: Obtain the temperature values of the temperature measurement points at each pressure relief valve of the lithium-ion battery module to be predicted; Input the obtained temperature values into a preset physical information-neural network model for prediction to obtain the temperature field on the surface of the lithium-ion battery module to be predicted; wherein, the preset physical information-neural network model includes: simplification and modeling of the heat transfer physical problem of the battery module, the basic neural network part and the physical information fusion part. The physical information fusion part adds the battery heat transfer differential equation and the battery boundary condition as physical information constraints during the neural network training process. The loss function in the neural network includes boundary loss, partial differential equation loss and observation data set loss.

2. The method for predicting the surface temperature field of a large-capacity energy storage lithium-ion battery module according to claim 1, wherein The simplification and modeling of the heat transfer physical problem of the battery module includes: For the convenience of modeling, the battery module is conceptually unfolded with the top surface as the central reference plane. Through this transformation, the temperature distribution on 11 surfaces represents the surface temperature of the battery module; the steady-state temperature distribution of the battery is controlled by the following equation: When solving the heat transfer problem of the battery module, first analyze the heat conduction mechanism inside a single battery and between adjacent batteries; through the layout of heat sources and the focused analysis of the heat conduction mechanism, simplify the calculation of the battery temperature field and achieve the best balance between accelerating the training process and calculation accuracy.

3. The prediction method for the temperature field of the large-capacity energy storage lithium-ion battery module according to claim 1, wherein The temperature field prediction based on physical information-neural network includes: In the physical information-neural network model, first estimate the initial heating power according to the approximate temperature distribution of the battery, and then train the weights and biases of the model based on the heat balance equation; for each heating unit, combine the heat conduction physical processes inside and between batteries to construct the loss function defined by equations (2)-(4): Loss bound = Loss bound_x + Loss bound_y + Loss bound_z (3) Among them, W obs , and W bound and represent the weight parameters of the observed data, partial differential equation, and boundary conditions respectively; Loss pde , Loss obs , Loss bound , Loss represent the partial differential equation loss, observed data loss, boundary condition loss, and total loss respectively; Loss bound_x , Loss bound_y , Loss bound_z represent the boundary losses in the x, y, and z directions respectively; among them, the thermal conductivities in the x, y, and z directions are set to 10.63 W / (m·K), 2.55 W / (m·K), and 7.74 W / (m·K) respectively, and the heat transfer coefficient between the battery surface and the air is 10 W / (m 2 ·K); by minimizing this loss function, the MLP network is trained to predict the battery temperature.

4. The prediction method for the temperature field of a large-capacity energy storage lithium-ion battery module according to claim 1, characterized in that, In the preset physical information-neural network model, the basic neural network part is constructed by multi-layer perceptrons; the physical information fusion part refers to the part where the loss function in the neural network is minimized during training.

5. The prediction method for the temperature field of the large-capacity energy storage lithium-ion battery module according to claim 4, characterized in that, During the construction of the multi-layer perceptron, use the open-source deep learning framework PyTorch; in the physical information-neural network model, the depth of the multi-layer perceptron MLP is 6, which consists of an input layer, four hidden layers and an output layer, and each layer contains 100 neurons; select the tanh function as the activation function; the initial learning rate is set to 1e-4, and the weights of the partial differential equation, boundary condition and observation data are set to 1, 1 and 1e-4 respectively.

6. The prediction method for the temperature field of a large-capacity energy storage lithium-ion battery according to claim 4, wherein The input of the physical information-neural network model is the temperature data measured by the limited pressure relief valve of the battery verified by experiments. At this time, the loss function is the sum of the boundary loss, partial differential equation loss and observation data set loss. Among them, the boundary condition loss term includes the heat transfer inside the battery module and between the module and the air. The approximate range of the temperature of some surfaces of the battery is added to the partial differential equation loss. The network parameters are trained by minimizing the Loss, and finally the optimal parameters of the battery temperature reconstruction model are obtained, and then the predicted temperature field parameter set is obtained.

7. Prediction system for temperature field of large-capacity energy storage lithium-ion battery module, characterized in that, Including: A data acquisition module that obtains the temperature values of the temperature measurement points at each pressure relief valve of the lithium-ion battery module to be predicted; The data processing module inputs the obtained temperature values into a preset physical information-neural network model for prediction to obtain the temperature field on the surface of the lithium-ion battery module to be predicted. Among them, the preset physical information-neural network model includes: simplification and modeling of the heat transfer physical problem of the battery module, the basic neural network part, and the physical information fusion part. The physical information fusion part adds the battery heat transfer differential equation and the battery boundary condition as physical information constraints during the neural network training process. The loss function in the neural network includes boundary loss, partial differential equation loss, and observed dataset loss.

8. The prediction system for the temperature field of a large-capacity energy storage lithium-ion battery according to claim 7, wherein In the data acquisition module, the temperature values of a few temperature measurement points of the lithium-ion battery module to be predicted are obtained, including: Based on the battery temperature measurement experimental platform, temperature measurement experiments are carried out on a three-series 280Ah battery module under different working conditions to obtain the temperature measurement values at the pressure relief valve on the surface of the battery module, which are used as the temperature values of a few temperature measurement points of the lithium-ion battery module to be predicted.

9. The prediction system for the temperature field of a large-capacity energy storage lithium-ion battery according to claim 7, wherein In the data processing module, the simplification and modeling of the heat transfer physical problem of the battery module include: For the convenience of modeling, the battery module is conceptually unfolded with the top surface as the central reference plane. Through this transformation, the temperature distribution on 11 surfaces represents the surface temperature of the battery module. The steady-state temperature distribution of the battery is controlled by the following equation: When solving the heat transfer problem of the battery module, the heat conduction mechanism inside a single battery and between adjacent batteries is analyzed preferentially. Through the arrangement of heat sources and the focused analysis of the heat conduction mechanism, the calculation of the battery temperature field is simplified, and the best balance is achieved between the acceleration of the training process and the calculation accuracy.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, it realizes the steps of the method for predicting the temperature field of a large-capacity energy storage lithium-ion battery module according to any one of claims 1-7.

Images