Electrolytic bath three-dimensional temperature field reconstruction method and system based on multi-physics field coupling

Abstract

The invention provides an electrolytic bath three-dimensional temperature field reconstruction method and system based on multi-physics field coupling, and belongs to the technical field of energy equipment thermal management, and the reconstruction method is applied to a reconstruction system comprising an acquisition module, a first processing module, a second processing module, a third processing module, a fourth processing module and a rendering module. The method specifically comprises the following steps: layering a three-dimensional model of the electrolytic cell according to a Z-axis direction, establishing a hierarchical database and setting a reference layer; generating initial temperature field distribution of the reference layer based on an anisotropic radial basis function interpolation algorithm; and establishing an electric-thermal-flow coupling equation, solving according to the hierarchical database, obtaining temperature field distribution of a reference layer, mapping the temperature field distribution to a three-dimensional space, correcting the temperature field distribution by combining interlayer temperature attenuation and a transverse thermal diffusion effect, obtaining target temperature field distribution, and carrying out visualization processing on the target temperature field distribution. Nonlinear temperature distribution of the electrolytic cell can be completely reproduced, and the accuracy of three-dimensional temperature field reconstruction of the electrolytic cell is guaranteed.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of energy equipment thermal management, in particular to an electrolytic cell three-dimensional temperature field reconstruction method and system based on multi-physical field coupling. BACKGROUND

[0002] Due to the influence of the closed structure of the electrolytic cell, it is not possible to implement large-scale deployment of sensors inside the electrolytic cell, and usually only a few discrete sensors can be used to identify the temperature distribution inside the electrolytic cell to implement subsequent online thermal management or fault diagnosis. However, this temperature distribution identification method relying on a few discrete sensors cannot capture the details of the temperature field distribution inside the electrolytic cell, and the lack of such thermal distribution details not only accelerates the proton membrane aging of the electrolytic cell due to abnormal flow channel heat accumulation, but also causes catalyst activity attenuation, and further increases the false alarm rate of the electrolytic cell, affecting the operation safety of the energy equipment such as hydrogen energy equipment and fuel cells where the electrolytic cell is located.

[0003] Under the condition of limited sensor configuration, the temperature field distribution of the electrolytic cell is usually reconstructed to capture the thermal distribution details inside the electrolytic cell to ensure the operation safety of the energy equipment where the electrolytic cell is located. However, as a closed three-dimensional layered equipment, the electrolytic cell contains multiple components such as bipolar plates, membrane electrodes, and flow channels, and there is strong coupling of multiple physical fields such as electricity, heat, flow, and chemical reactions inside the electrolytic cell. Under the action of multi-physical field coupling, the electrolytic cell inside will show nonlinear temperature distribution characteristics. However, the current mainstream electrolytic cell temperature field reconstruction methods, such as linear interpolation methods based on uniform grids, traditional inversion models, and Kriging interpolation algorithms, cannot adapt to the high-precision temperature field distribution reconstruction requirements of the electrolytic cell in such a three-dimensional scenario, and easily ignore the nonlinear temperature distribution characteristics shown by the electrolytic cell inside, resulting in low accuracy of temperature distribution identification inside the electrolytic cell. SUMMARY

[0004] The present application aims to overcome the shortcomings in the prior art that easily ignore the nonlinear temperature distribution characteristics shown by the electrolytic cell inside under the action of multi-physical field coupling, and provide an electrolytic cell three-dimensional temperature field reconstruction method and system based on multi-physical field coupling, which focuses on the temperature field of different components through Z-axis layered modeling, and combines anisotropic radial basis function interpolation to adapt to the complex flow channel heat transfer characteristics, and further solves the electro-thermal-flow coupling equation based on the sampling data to incorporate the multi-physical field mechanism, reduces the flow channel temperature rise deviation and key area temperature estimation error, and then maps the reference layer temperature field to the three-dimensional space in layers, and combines the interlayer temperature attenuation and transverse thermal diffusion effect correction to completely reproduce the nonlinear temperature distribution, and ensures the accuracy of the electrolytic cell three-dimensional temperature field reconstruction.

[0005] The present application is achieved by the following technical solutions: The electrolytic cell three-dimensional temperature field reconstruction method based on multi-physical field coupling comprises: The three-dimensional model of the electrolytic cell is layered according to the Z-axis direction, a layered database is established according to the layered results, and a reference layer is set; Based on the anisotropic radial basis function interpolation algorithm, the initial temperature field distribution of the reference layer is generated; According to the initial temperature field of the reference layer, an electro-thermal-fluid coupling equation is established, the layered database is sampled, and the electro-thermal-fluid coupling equation is solved according to the sampling data to obtain the temperature field distribution of the reference layer; The temperature field distribution of the reference layer is layered and mapped to the three-dimensional space along the Z-axis direction, and the temperature field distribution of the reference layer is corrected in combination with the interlayer temperature attenuation and the transverse heat diffusion effect to obtain the target temperature field distribution of the reference layer; The target temperature field distribution is rendered to obtain a visual image of the three-dimensional temperature field of the electrolytic cell.

[0006] By Z-axis layered modeling, the temperature field of different components is focused, avoiding the loss of local details caused by overall modeling, and combining the anisotropic radial basis function interpolation to adapt to the heat transfer anisotropy caused by the current density gradient of the bipolar plate under the complex flow channel, the temperature rise prediction deviation of the serpentine flow channel and other areas is reduced. Based on the sampling data, the electro-thermal-fluid coupling equation is solved to integrate the multi-physical field mechanism of electrochemical reaction, heat conduction and fluid flow, further reducing the temperature estimation error of the boundary heat transfer enhancement area and the membrane electrode interface. At the same time, the temperature field of the reference layer is layered and mapped to the three-dimensional space, and the temperature field is corrected in combination with the interlayer temperature attenuation and the transverse heat diffusion effect to completely reproduce the overall nonlinear temperature distribution of the layered structure, ensuring the accuracy of the electrolytic cell three-dimensional temperature field reconstruction.

[0007] Further, the three-dimensional model of the electrolytic cell is layered according to the Z-axis direction, a layered database is established according to the layered results, and a reference layer is set, comprising: Discretize the three-dimensional model of the electrolytic cell into multiple two-dimensional sections along the Z-axis direction; According to the curvature analysis and normal vector clustering algorithm, the multiple two-dimensional sections are adaptively layered in combination with the vertex curvature adaptive layering strategy; Extract the three-dimensional coordinate data of each layer, establish a layered database, and select a reference layer from the layered.

[0008] Further, the initial temperature field distribution of the reference layer is generated based on the anisotropic radial basis function interpolation algorithm, comprising: Obtain the main diffusion axis of the flow channel cross-section direction through principal component analysis, and establish an anisotropic distance metric function; Based on the anisotropic distance metric function, the data in the layered database is spatially interpolated, and the initial temperature field distribution of the reference layer is generated in combination with the regularization parameter optimization.

[0009] Further, the sampling of the layered database comprises: dividing the reference layer into four quadrants, calculating the average temperature gradient in each quadrant, and assigning a weight to each quadrant according to the average temperature gradient; allocating the number of sampling points to each quadrant based on the corresponding weight, and generating additional sampling points in combination with a boundary protection mechanism; sampling the layered database according to the corresponding number of sampling points and the generated additional sampling points to obtain sampling data.

[0010] Further, the solving of the electro-thermal-fluid coupling equation according to the sampling data to obtain the temperature field distribution of the reference layer comprises: discretizing the electro-thermal-fluid coupling equation by finite volume method and introducing an artificial diffusion term; setting a soft constraint boundary condition, inputting the sampling data as a boundary, and iteratively solving the discretized electro-thermal-fluid coupling equation with the artificial diffusion term to obtain the temperature field distribution of the reference layer.

[0011] Further, the temperature field distribution of the reference layer is mapped to a three-dimensional space along the Z-axis direction, and the temperature field distribution of the reference layer is corrected in combination with the interlayer temperature attenuation and lateral heat diffusion effect to obtain the target temperature field distribution of the reference layer, comprising: calculating the temperature attenuation coefficient of the temperature field distribution of the reference layer along the Z-axis, and establishing a spatial mapping relationship between the non-reference layer vertex and the reference layer according to the fast nearest neighbor search algorithm; correcting the temperature field distribution of the reference layer according to the temperature attenuation coefficient and the lateral heat diffusion effect to obtain the target temperature layer distribution of the reference layer.

[0012] Further, the rendering of the target temperature field distribution to obtain the visualized image of the three-dimensional temperature field of the electrolytic cell comprises: rendering the target temperature field distribution of the reference layer based on a multi-modal rendering engine to obtain the surface isotherm cloud chart and three-dimensional volume rendering chart of the electrolytic cell.

[0013] The electrolytic cell three-dimensional temperature field reconstruction system based on multi-physical field coupling is used to execute any one of the electrolytic cell three-dimensional temperature field reconstruction methods based on multi-physical field coupling, comprising: an acquisition module for acquiring a three-dimensional model of an electrolytic cell and layering according to the Z-axis direction, establishing a layered database and setting a reference layer according to the layering result; a first processing module for generating an initial temperature field distribution of the reference layer based on an anisotropic radial basis function interpolation algorithm; a second processing module for sampling the layered database to obtain corresponding sampling data; The third processing module is configured to input the sampling data into an electro-thermal-fluid coupling equation, solve an initial temperature field distribution of the reference layer, and obtain the temperature field distribution of the reference layer. The fourth processing module is configured to map the temperature field distribution of the reference layer to a three-dimensional space along a Z-axis direction, correct the temperature field distribution of the reference layer by combining interlayer temperature attenuation and transverse thermal diffusion effects, and obtain a target temperature field distribution of the reference layer. The rendering module is configured to render the target temperature field distribution of the reference layer to obtain a visual image of the three-dimensional temperature field of the electrolytic cell.

[0014] An electronic device includes a memory, a processor, and a computer program stored on the memory and executable on the processor, and the processor implements the method for reconstructing a three-dimensional temperature field of an electrolytic cell based on multi-physical field coupling according to any one of the embodiments.

[0015] A non-transitory computer-readable storage medium has a computer program stored thereon, and the computer program is executed by a processor to implement the method for reconstructing a three-dimensional temperature field of an electrolytic cell based on multi-physical field coupling according to any one of the embodiments.

[0016] The present application has the following advantages: (1) The Z-axis layering modeling focuses on the temperature field of different components, avoids the loss of local details caused by overall modeling, and combines anisotropic radial basis function interpolation to adapt to the heat transfer anisotropy caused by the current density gradient of the bipolar plate under complex flow channels, thereby reducing the temperature rise prediction deviation in areas such as serpentine flow channels. Based on the sampling data, the electro-thermal-fluid coupling equation is solved to integrate the multi-physical field mechanism of electrochemical reaction, heat conduction, and fluid flow, further reducing the temperature estimation error of the boundary heat transfer enhancement area and the membrane electrode interface. Meanwhile, the temperature field of the reference layer is mapped to a three-dimensional space in layers, and the temperature field is corrected by combining interlayer temperature attenuation and transverse thermal diffusion effects, so as to completely reproduce the overall nonlinear temperature distribution of the layered structure and ensure the accuracy of the reconstruction of the three-dimensional temperature field of the electrolytic cell.

[0017] (2) The Z-axis discretization combined with curvature analysis, normal vector clustering, and vertex curvature adaptive layering strategy matches the structural differences of different components of the electrolytic cell, avoids the loss of local structural temperature information caused by traditional equal layering, and further constructs a corresponding layering database and sets a reference layer, so as to more accurately simulate the temperature distribution of different levels and effectively improve the accuracy of the reconstruction of the temperature field distribution of the electrolytic cell.

[0018] (3) The main diffusion axis of the flow channel cross section is determined by principal component analysis, and an anisotropic distance metric function is constructed to make the spatial interpolation more consistent with the anisotropic characteristics of heat transfer in the electrolytic cell. Combined with the optimization of the regularization parameter, the interpolation error in the complex flow channel area is effectively reduced, and the accuracy of the initial temperature field distribution of the reference layer is improved. Further combined with the data sampling of the layered database, the electro-thermal-flow coupling equation constructed according to the initial temperature field distribution is solved. In the solving process, the stability of the solution is improved by discretization and artificial diffusion term introduction through the finite volume method, and combined with the soft constraint boundary condition, the solution result is closer to the real running state of the electrolytic cell, and the temperature estimation deviation of the key parts such as the boundary reinforced heat transfer area is reduced, and the accuracy of the obtained reference layer temperature field distribution is ensured.

[0019] (4) The spatial mapping relationship established by temperature decay coefficient calculation and fast nearest neighbor search realizes the extension of the reference layer temperature field to three-dimensional space, and the reference layer temperature field is corrected by combining interlayer temperature decay and lateral heat diffusion effect. The error caused by ignoring interlayer heat transfer can be effectively compensated, and the nonlinear temperature distribution in three-dimensional space can be completely reproduced. BRIEF DESCRIPTION OF DRAWINGS

[0020] Figure 1 is a flowchart of the present application; Figure 2 is a structure schematic diagram of a three-dimensional model before layering of an embodiment of the present application; Figure 3 is a structure schematic diagram of a three-dimensional model after layering of an embodiment of the present application; Figure 4 is a structure schematic diagram of a three-dimensional model after final layering of an embodiment of the present application; Figure 5 is a flowchart of visualization of a three-dimensional temperature field of an electrolytic cell of an embodiment of the present application; Figure 6 is a structure schematic diagram of a three-dimensional temperature field reconstruction system of an electrolytic cell of an embodiment of the present application; Figure 7 is a structure schematic diagram of an electronic device of an embodiment of the present application.

[0021] Among them: 1, the acquisition module, 2, the first processing module, 3, the second processing module, 4, the third processing module, 5, the fourth processing module, 6, the rendering module, 7, the electronic device, 71, the memory, 72, the processor. DETAILED DESCRIPTION

[0022] The present application is further described below in conjunction with the drawings and embodiments.

[0023] Embodiment: The purpose of the present application is achieved by the following technical solutions: With reference to the drawings and the embodiments of the present application, the technical solutions in the embodiments of the present application will be clearly described. Obviously, the described embodiments are only some of the embodiments of the present application, but not all of the embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those of ordinary skill in the art are within the scope of the present application.

[0024] The terms "first", "second", and the like in the specification and claims of the present application are used to distinguish similar objects, and are not used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged under appropriate circumstances, so that the embodiments of the present application can be implemented in an order other than those illustrated or described herein, and the objects distinguished by "first", "second", etc. are generally a category, and are not limited to the number of objects, for example, the first object can be one or more. In addition, "and / or" in the specification and claims indicates at least one of the connected objects, and the character " / ", generally indicates that the front and rear associated objects are in an "or" relationship.

[0025] The method for reconstructing a three-dimensional temperature field of an electrolytic cell based on multi-physical field coupling, the device for reconstructing a three-dimensional temperature field of an electrolytic cell based on multi-physical field coupling, the electronic device, and the readable storage medium provided by the embodiments of the present application will be described in detail below with reference to the drawings and specific embodiments and application scenarios.

[0026] The method for reconstructing a three-dimensional temperature field of an electrolytic cell based on multi-physical field coupling can be applied to a terminal, and can be executed by hardware or software in the terminal.

[0027] The terminal includes, but is not limited to, a portable communication device such as a mobile phone or a tablet computer having a touch-sensitive surface (e.g., a touchscreen display and / or a touchpad). It should also be understood that, in some embodiments, the terminal can not be a portable communication device, but a desktop computer having a touch-sensitive surface (e.g., a touchscreen display and / or a touchpad).

[0028] In the following various embodiments, a terminal including a display and a touch-sensitive surface is described. However, it should be understood that the terminal can include one or more other physical user interface devices such as a physical keyboard, a mouse, and a joystick.

[0029] The method for reconstructing a three-dimensional temperature field of an electrolytic cell based on multi-physical field coupling provided in the embodiments of the present application can be executed by an electronic device or a functional module or functional entity of the electronic device capable of realizing the method for reconstructing a three-dimensional temperature field of an electrolytic cell based on multi-physical field coupling. The electronic device mentioned in the embodiments of the present application includes but is not limited to a mobile phone, a tablet computer, a computer, a camera, a wearable device, and the like. The method for reconstructing a three-dimensional temperature field of an electrolytic cell based on multi-physical field coupling provided in the embodiments of the present application will be described below by taking an electronic device as an example.

[0030] The method for reconstructing a three-dimensional temperature field of an electrolytic cell based on multi-physical field coupling, as shown in Figure 1 includes the following steps. Layering a three-dimensional model of the electrolytic cell in the Z-axis direction, establishing a layered database according to the layering result, and setting a reference layer; Generating an initial temperature field distribution of the reference layer based on an anisotropic radial basis function interpolation algorithm; Establishing an electro-thermal-fluid coupling equation according to the initial temperature field of the reference layer, sampling the layered database, solving the electro-thermal-fluid coupling equation according to the sampling data, and obtaining the temperature field distribution of the reference layer; Layering and mapping the temperature field distribution of the reference layer to a three-dimensional space along the Z-axis direction, correcting the temperature field distribution of the reference layer in combination with interlayer temperature attenuation and lateral heat diffusion effect, and obtaining a target temperature field distribution of the reference layer; Rendering the target temperature field distribution to obtain a visualized image of the three-dimensional temperature field of the electrolytic cell.

[0031] The electrolytic cell is stacked by multiple components such as bipolar plates, membrane electrodes, and flow channels along the Z-axis direction. The material properties, structural forms, and heat transfer characteristics of different components are significantly different, resulting in strong layered heterogeneity of the temperature field in the Z-axis direction. If the three-dimensional space is regarded as a unified whole to construct the three-dimensional temperature field of the electrolytic cell, the interlayer differences are easily ignored, resulting in smoothing of local details such as temperature changes at the contact interface between the membrane electrode and the bipolar plate, and the real heat distribution cannot be reflected. Moreover, the three-dimensional model data is large, and the calculation amount is large if direct global calculation is performed, and the reconstruction efficiency is low.

[0032] Therefore, by layering and decomposing for fine modeling, the three-dimensional problem is divided into layered two-dimensional and interlayer correlation, the calculation amount is reduced, and the reconstruction amount is improved.

[0033] The layering of the three-dimensional model of the electrolytic cell in the Z-axis direction, the establishment of a layered database according to the layering result, and the setting of a reference layer include the following steps. Discretizing the three-dimensional model of the electrolytic cell into multiple two-dimensional sections along the Z-axis direction; According to curvature analysis and normal vector clustering algorithm, combined with vertex curvature adaptive layering strategy, the multi-layer two-dimensional section is adaptively layered; The three-dimensional coordinate data of each layer is extracted, a layered database is established, and a reference layer is selected from the layers.

[0034] Specifically, the corresponding three-dimensional model is determined by extracting the STL (Stereolithography) model file of the electrolytic cell. The STL model file records the surface geometric topology information of the components such as the bipolar plate, membrane electrode and flow channel of the electrolytic cell.

[0035] Then, the Z-axis layering algorithm is used to identify the distribution range of the Z-axis coordinates of the vertices, and the component stacking order of the electrolytic cell along the height direction is determined, so as to obtain the topology structure of the three-dimensional model. Then, the initial discrete interval is set along the Z-axis, and the complete three-dimensional model is cut into a series of two-dimensional sections parallel to the XY plane. The structure diagram of the three-dimensional model before layering is shown in Figure 2 The structure diagram of the three-dimensional model after layering is shown in Figure 3 The expression of the layer spacing during layering is: .

[0036] The calculation formula of the optimal number of layers is: ; Wherein, is the complexity penalty factor, the default value is 0.01, the feature point data provided by a small amount of sensors is initialized, is the optimal number of layers, MAE is the average absolute error of the temperature field reconstruction, is the layer height, that is, the thickness of a single layer.

[0037] At the same time, the three-dimensional vertex coordinate data is extracted to convert the discrete triangular facet information in the STL model file into structured coordinate data arranged in order along the Z-axis direction.

[0038] Then, the curvature analysis is used to identify the structural complexity of each two-dimensional section. The greater the geometric shape changes, the higher the vertex curvature value. Then, the normal vector clustering algorithm is used to classify the regions with similar normal vector directions in each section, so as to further clarify the partition of flat and complex structures.

[0039] Based on the results of curvature analysis and normal vector clustering algorithm, the two-dimensional sections obtained are adaptively layered by the vertex curvature adaptive layering strategy. The layering interval of the regions with high curvature and complex structure is reduced, the number of two-dimensional sections is increased to retain details, and the layering interval of the regions with low curvature and flat structure is increased to reduce invalid sections. The structure diagram of the finally constructed three-dimensional model after final layering is shown in Figure 4 .

[0040] Finally, all vertex three-dimensional coordinates of each two-dimensional section are extracted one by one, and the component attributes of the corresponding layer are associated, and the data is stored in order according to the layer number to build a layered database. In addition to the three-dimensional coordinate data in the layered database, the corresponding feature point distribution and direction amount direction are further recorded to provide a geometric reference for subsequent interpolation. On this basis, a reference layer is selected from all layers. In the selection process, the layer containing the core reaction and heat transfer area of the electrolytic cell is preferentially selected.

[0041] Although the layered database has structured three-dimensional coordinate data and component attributes, it lacks corresponding temperature distribution information, and the reference layer contains flow channels, membrane electrodes and other components. The flow channel area presents anisotropic characteristics of strong heat transfer along the flow channel direction and weak heat transfer perpendicular to the flow channel direction due to fluid flow. Moreover, the number of discrete sensors installed in the electrolytic cell is small, and traditional interpolation methods cannot generate an initial temperature field that fits the true thermal state of the reference layer. Therefore, the embodiment generates the initial temperature field distribution of the reference layer using an anisotropic radial basis function interpolation algorithm.

[0042] Specifically, the anisotropic radial basis function interpolation algorithm generates the initial temperature field distribution of the reference layer, including: Obtaining the main diffusion axis of the flow channel cross-section direction through principal component analysis, and establishing an anisotropic distance measurement function; Based on the anisotropic distance measurement function, spatial interpolation is performed on each data in the layered database, and the initial temperature field distribution of the reference layer is generated by combining the regularization parameter optimization.

[0043] First, the key direction of dominant temperature diffusion, i.e., the main diffusion axis, is extracted from the geometric and heat transfer data of the flow channel of the reference layer through principal component analysis. Specifically, the three-dimensional coordinate data of the flow channel cross-section of the reference layer in the layered database can be used in combination with the fluid flow direction and heat transfer data in the flow channel to identify the direction with the largest data variance through principal component analysis, and determine it as the main diffusion axis of the flow channel cross-section.

[0044] Specifically, the calculation formula of the main diffusion axis through principal component analysis is: ; Where Cov is the covariance matrix, N is the number of sample points, is the two-dimensional coordinate of the kth sample point, is the sample point mean vector.

[0045] According to the main diffusion axis, an anisotropic distance weight matrix is calculated wherein, , is the PCA eigenvalue, and then the space metric of the radial basis function kernel function is adjusted, and the expression of the modified radial basis function kernel function is: ; wherein, is the anisotropic radial basis kernel function, is the shape parameter, i.e. the kernel function width, is the anisotropic distance weight matrix.

[0046] Then, the shape parameter of the Gaussian kernel function is dynamically adjusted based on the nearest neighbor statistical result, and the expression of the adjustment formula is as follows: ; wherein, is the shape parameter of the radial basis function kernel function, and N is the number of sampling points of the reference layer, and are the coordinates of the feature point i and the feature point j.

[0047] On this basis, each data in the layered database is spatially interpolated, and in the interpolation process, with the limited discrete temperature measurement points as the core, the heat transfer correlation degree of each unknown temperature point in the layered database and the temperature measurement point is calculated according to the anisotropic distance metric function composed of the anisotropic distance weight matrix and the radial basis function kernel function, and the temperature value of the unknown point is calculated by the correlation degree weighting.

[0048] In the interpolation process, a Tikhonov regularization term is also introduced in the coefficient matrix of the radial basis function kernel function, and the regularization parameter is optimized and selected by the L-curve criterion, so as to improve the stability and accuracy of the interpolation by optimizing the regularization parameter.

[0049] The matrix element in the coefficient matrix of the radial basis function kernel function .

[0050] The expression of the Tikhonov regularization term is: ; wherein, L is a second-order difference matrix, selected by L-curve method, c is the coefficient vector to be solved, A is the design matrix, is the observed temperature vector.

[0051] Finally, according to the generated temperature values of each point, the initial temperature field distribution of the reference layer is obtained. The finally generated initial temperature field distribution of the reference layer not only contains high-resolution temperature data, but also includes temperature gradient distribution diagram.

[0052] ​Considering the large amount of data in the hierarchical database, in order to improve the subsequent calculation efficiency, the hierarchical database is further sampled. The generated initial temperature field has shown the temperature distribution trend of the reference layer, especially through anisotropic interpolation, the high gradient area characteristics such as flow channel corner heat accumulation and membrane electrode interface temperature change, and the low gradient area characteristics such as the plate plane, and the gradient difference directly determines the influence weight of different areas on the reconstruction accuracy of the temperature field. The high gradient area is the core area of the thermal risk and the sensitive area of the subsequent coupled equation solving. If the sampling is insufficient, the key thermal characteristics are easy to be lost, and the temperature distribution in the low gradient area is stable. Excessive sampling will cause data redundancy and waste of computing resources. Therefore, the four-quadrant weighted sampling algorithm is used to realize the data sampling processing of the hierarchical database.

[0053] Specifically, the sampling of the hierarchical database comprises: The reference layer is divided into four quadrants, the average temperature gradient in each quadrant is calculated, and the weight of each quadrant is allocated according to the average temperature gradient; The number of sampling points of each quadrant is allocated based on the corresponding weight, and additional sampling points are generated combined with the boundary protection mechanism; According to the number of corresponding sampling points and the generated additional sampling points, the hierarchical database is sampled to obtain sampling data.

[0054] Firstly, the reference layer is divided into four quadrants along the X-axis and Y-axis middle line to more accurately locate the gradient difference of different areas. Then the average temperature gradient in each quadrant is calculated Then, the weight is allocated according to the average temperature gradient. The higher the average gradient value of the quadrant, the more intense the temperature change in the region, the more prominent the thermal risk, and the greater the influence on the accuracy of the subsequent coupled equation solving. Therefore, a higher weight is given, and vice versa, a low weight is given to the quadrant with a low average gradient value.

[0055] The calculation formula of the average temperature gradient is: ; Wherein, is the average temperature gradient amplitude of the qth quadrant, is the number of data points in the qth quadrant.

[0056] According to the corresponding sampling point number allocated according to the quadrant weight, the sampling density of the high gradient area and the sampling density of the low gradient area satisfy: ; Wherein, is the sampling density of the high gradient area, is the initial reference sampling density, is the temperature gradient of the qth quadrant, is the sampling density of the low gradient area.

[0057] The expression of the sampling point number allocation is: ; wherein, =0.1 is a smoothing factor for preventing division by zero, is the number of sampling points allocated to the qth quadrant, is the total number of sampling points, is the average temperature gradient amplitude of the qth quadrant, is a smoothing adjustment parameter.

[0058] Meanwhile, additional sampling points are generated in combination with the boundary protection mechanism, which are arranged at the boundary lines of each quadrant, the physical boundaries of the reference layer, and the boundary regions of the gradient mutation, which neither affect the sampling density of the high gradient region nor provide clear boundary temperature input for the coupled equation, thereby avoiding unreasonable fluctuations in the boundary temperature when solving.

[0059] The arrangement interval of the additional sampling points needs to meet: ; wherein, is a boundary attenuation coefficient, and the value range is 0.2-0.5.

[0060] At this time, the reference layer only has an initial temperature field distribution, which is a preliminary temperature framework generated by anisotropic interpolation, and has not been integrated into the coupling effect of the internal electric, thermal, and flow multi-physical fields of the electrolytic cell, which deviates from the real running state and cannot be directly used as the final temperature result of the reference layer.

[0061] Therefore, the control equations of the electric, thermal, and flow three physical fields are first established, the electric domain equation describes the current density distribution and ohmic heat, the thermal domain equation describes the temperature change and heat transfer, and the flow domain equation describes the fluid flow state, and then the control equations of the three physical fields are coupled to describe the coupling effect of the internal multi-physical fields of the electrolytic cell.

[0062] The expression of the electric-thermal-flow coupled equation is: ; wherein, k is the equivalent thermal conductivity coefficient corrected by the membrane thickness, Q is the joule heat source term calculated based on the current density, is the fluid density, is the specific heat capacity, T is the temperature field, and t is the time variable, is the flow velocity vector field.

[0063] By introducing the key physical quantities such as the equivalent thermal conductivity corrected by the membrane thickness, the Joule heat source term, the fluid density, the specific heat capacity and the flow velocity vector field, and combining the change of the temperature field distribution with time and space, the complex electro-thermal-fluid coupling process is accurately simulated, and the accurate prediction of the temperature field can be realized by solving the electro-thermal-fluid equation combined with the sampling data.

[0064] Specifically, the temperature field distribution of the reference layer is obtained by solving the electro-thermal-fluid coupling equation according to the sampling data, including: The electro-thermal-fluid coupling equation is discretized by the finite volume method, and an artificial diffusion term is introduced. The soft constraint boundary condition is set, the sampling data is taken as the boundary input, and the electro-thermal-fluid coupling equation after discretization and introduction of the artificial diffusion term is iteratively solved to obtain the temperature field distribution of the reference layer.

[0065] The electric, thermal and flow fields of the reference layer of the electrolytic cell are continuous distribution physical fields, and the control equation exists in the form of partial differential equation, which cannot be directly solved. Therefore, the space of the reference layer is divided into a large number of small control bodies by the finite volume method, and each control body is taken as a basic calculation unit to integrate the partial differential equation in the control body, so that the complex partial differential equation group is converted into an algebraic equation group containing thousands to tens of thousands of unknowns, making the electro-thermal-fluid equation solvable.

[0066] Specifically, the expression of the finite volume method is: ; Wherein, is the electrical conductivity, and V is the potential fraction. For each control body Integrate, and the convection term is discretized by the second-order upwind format.

[0067] And in the flow channel corner, the membrane electrode and the bipolar plate contact area, etc., because of the sudden change of physical parameters, the discrete algebraic equation group may appear sharp fluctuation of solution, therefore, in order to avoid the risk of numerical oscillation after discretization, the artificial diffusion term is introduced to ensure that the flow channel heat accumulation, membrane electrode interface temperature change and other key gradient characteristics are not covered while suppressing oscillation, and the stability and result accuracy of the solution are balanced.

[0068] The expression of the artificial diffusion term is: ; Wherein, is the stability factor, and the value range is 0.1-0.3, is the spatial discrete step.

[0069] And the stability factor of the artificial diffusion term can be adaptively adjusted according to the local grid size and temperature gradient obtained after discretization.

[0070] There are clear boundary conditions when the electrolytic cell is running, such as the inlet coolant temperature of the flow channel, the heat dissipation coefficient of the edge of the plate, etc. If the boundary conditions are unknown when solving, the equation will have an unconstrained solution, which is seriously inconsistent with the actual working condition. Therefore, a soft constraint boundary condition is set to allow the sampled data to adapt to the equation logic within a small range, which not only ensures that the boundary temperature is close to the measured data, but also avoids the equation having no solution due to the small error of the sampled data. The sampled data of the four-quadrant sampling is classified according to the physical field, and the temperature data corresponds to the thermal domain boundary, the flow rate data corresponds to the flow domain boundary, and the current data corresponds to the electric domain boundary, which ensures that the solving result conforms to the physical logic of the actual operation of the electrolytic cell.

[0071] Based on the above processed equation, an iterative solution is performed. First, a set of initial values is assumed, including the average temperature of the reference layer, the average flow rate of the flow channel, the average current density and other parameters, which are substituted into the discrete algebraic equation set to calculate the temperature, flow rate and current density of each control body. Then, the deviation of the calculation result from the sampling data is compared, and the initial value is adjusted according to the deviation result, and then substituted into the equation for calculation. Repeat the process of assumption, calculation, deviation and correction until the deviation of all sampling points is less than the set threshold value, at which time the equation converges, and the obtained temperature distribution is the temperature field distribution of the reference layer. And in the solving process, the mass conservation error is also monitored in real time, and when the residual error exceeds the threshold value, the feedback mechanism is triggered to add sampling points in a specific area, forming a closed-loop optimization system, and finally outputting the temperature field distribution that satisfies the law of conservation of mass.

[0072] Since the obtained temperature field distribution of the reference layer is still essentially a single two-dimensional cross-sectional temperature distribution, it does not cover the full spatial structure of the electrolytic cell along the Z-axis, and ignores the interlayer heat transfer effect, which still deviates from the actual situation. Therefore, further through three-dimensional mapping and effect correction, the two-dimensional temperature information of the reference layer is expanded to three-dimensional temperature information including interlayer heat interaction.

[0073] Among them, the temperature field distribution of the reference layer is layered along the Z-axis direction to the three-dimensional space, and the temperature field distribution of the reference layer is corrected by combining the interlayer temperature attenuation and the lateral heat diffusion effect, to obtain the target temperature field distribution of the reference layer, which includes: The temperature attenuation coefficient of the temperature field distribution of the reference layer along the Z-axis is calculated, and the spatial mapping relationship between the non-reference layer vertex and the reference layer is established according to the fast nearest neighbor search algorithm; The temperature field distribution of the reference layer is corrected according to the temperature attenuation coefficient and the lateral heat diffusion effect, to obtain the target temperature layer distribution of the reference layer.

[0074] The calculation formula of the temperature attenuation coefficient is: ; Where L is the characteristic attenuation length, and z is the height difference between the current layer and the reference layer.

[0075] The layers of the electrolytic cell are stacked along the Z-axis. There are many structurally similar regions in the projection of the non-reference layer and the reference layer onto the XY plane. However, due to the difference in the thickness of the components between the layers, the Z coordinates of the vertices are different. Direct matching is prone to spatial misalignment. Therefore, the spatial mapping relationship between the vertices of the non-reference layer and the reference layer is established by the fast nearest neighbor search algorithm KDTree.

[0076] Specifically, when establishing the spatial mapping relationship between non-reference layer vertices and reference layer vertices using the fast nearest neighbor search algorithm, a KD-tree index structure is first constructed using the three-dimensional coordinates of all vertices in the reference layer. Then, each vertex in the non-reference layer is traversed, and the projection of the vertex onto the XY plane is used as the search target. The vertex with the closest coordinates and consistent structural properties is quickly found in the reference layer KD-tree, and the two are then mapped.

[0077] Based on the established spatial mapping relationship, for each vertex of the non-reference layer... Based on its nearest neighbor in the reference layer Calculate its mapped temperature using the following formula: ; in, This is a correction amount based on inter-layer gradient propagation. Temperature at the vertex of the non-reference layer. The temperature of the nearest neighbor point of the reference layer. This is the temperature decay coefficient.

[0078] Lateral thermal diffusion refers to the transfer of heat within the XY plane, and its correction mainly involves the thermal diffusion equation and the material's thermal diffusivity. When correcting the temperature field distribution of the reference layer based on the lateral thermal diffusion effect, a certain neighborhood is first defined within the XY plane for each vertex of both the reference and non-reference layers. Then, the average temperature of all vertices within the neighborhood is calculated. Combining the thermal diffusivity with a preset time step, the correction value for the vertex temperature is determined.

[0079] The formula for calculating the correction value is: ; in, This is a correction value for the vertex temperature of the non-reference layer. This is the initial value of the vertex temperature, i.e., the calculated mapped temperature. Where is the thermal diffusivity, For time step, The average temperature of all vertices in the neighborhood. is the neighborhood radius.

[0080] By calculating the correction values ​​of the mapping temperature and the vertex temperature, the temperature field distribution of the reference layer is corrected, thereby obtaining the target temperature field distribution of the reference layer, that is, the global three-dimensional temperature field distribution of the electrolytic cell.

[0081] After determining the target temperature field distribution, it is visualized. The process of rendering the target temperature field distribution to obtain a visualized image of the three-dimensional temperature field of the electrolytic cell includes: The target temperature field distribution of the reference layer is rendered using a multimodal rendering engine to obtain the surface isotherm cloud map and three-dimensional volume rendering map of the electrolytic cell.

[0082] Surface cloud maps and volume fields are output simultaneously using multimodal rendering technology, and temperature gradient perception is enhanced by ray casting optimization. Figure 5 This is a visual flowchart provided in this embodiment, such as... Figure 5 As shown, a multimodal rendering engine is used to visualize the three-dimensional temperature field, simultaneously generating surface isotherm cloud maps and three-dimensional volume rendering maps. The temperature gradient display is optimized through a ray casting algorithm, and interactive scalar bars and multi-plane cut views are dynamically generated.

[0083] Specifically, isothermal surfaces are extracted using the Marching Cubes algorithm, and the surface topology is optimized using Laplacian smoothing to generate surface isothermal cloud maps and 3D positive rendering maps.

[0084] A transparency transfer function is defined to optimize volumetric rendering. The expression for the transparency transfer function is: ; in, This is a temperature-dependent transparency function, where T is the current temperature value.

[0085] At the same time, the color mapping range is automatically adjusted according to the real-time temperature range, and key temperature thresholds are marked in the corresponding cloud map and rendering map.

[0086] Another aspect of this embodiment also provides a three-dimensional temperature field reconstruction system for an electrolytic cell based on multi-physics coupling, such as... Figure 6 As shown, it includes: Module 1 is used to acquire the three-dimensional model of the electrolytic cell, and to divide it into layers according to the Z-axis direction. Based on the layering results, a layered database is established and a reference layer is set. The first processing module 2 is used to generate the initial temperature field distribution of the reference layer based on the anisotropic radial basis function interpolation algorithm. The second processing module 3 is used to sample the hierarchical database and obtain corresponding sampled data; The third processing module 4 is configured to input the sampling data into an electro-thermal-flow coupling equation, solve an initial temperature field distribution of the reference layer, and obtain the temperature field distribution of the reference layer. The fourth processing module 5 is configured to map the temperature field distribution of the reference layer to a three-dimensional space in a Z-axis direction, correct the temperature field distribution of the reference layer by combining interlayer temperature attenuation and transverse thermal diffusion effects, and obtain a target temperature field distribution of the reference layer. The rendering module 6 is configured to render the target temperature field distribution of the reference layer to obtain a visual image of the three-dimensional temperature field of the electrolytic cell.

[0087] The acquisition module, the first processing module, the second processing module, the third processing module, the fourth processing module, and the rendering module are data processing elements with corresponding processing algorithms, such as computers and microprocessors, and are multi-thread parallel computing architectures with built-in four-quadrant dynamic allocation algorithms and boundary guarantee point generation logic, developed based on a PyVista framework, supporting temperature field multi-modal rendering and interactive analysis, and providing real-time data communication interfaces with external sensors and databases.

[0088] As shown in Figure 7 The electronic device 7 includes a memory 71, a processor 72, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements each process of the above-mentioned multi-physical field coupling-based electrolytic cell three-dimensional temperature field reconstruction method embodiment and achieves the same technical effects. To avoid repetition, details are not repeated here.

[0089] It should be noted that the electronic device in the embodiments of the present application includes the above-mentioned mobile electronic device and non-mobile electronic device.

[0090] The embodiments of the present application also provide a non-transitory computer readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, it implements each process of the above-mentioned multi-physical field coupling-based electrolytic cell three-dimensional temperature field reconstruction method embodiment and achieves the same technical effects. To avoid repetition, details are not repeated here.

[0091] The processor is the processor in the electronic device in the above-mentioned embodiments. The readable storage medium includes a computer readable storage medium, such as a computer readable memory (Read-Only Memory, ROM), a random access memory (Random Access Memory, RAM), a magnetic disk or an optical disk, etc.

[0092] The embodiments of the present application also provide a computer program product, including a computer program, which is executed by a processor to implement the above-mentioned multi-physical field coupling-based electrolytic cell three-dimensional temperature field reconstruction method.

[0093] The processor is a processor in the electronic device in the above-mentioned embodiments. The readable storage medium includes a computer readable storage medium, such as a computer read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0094] The embodiments of the present application further provide a chip, which comprises a processor and a communication interface, the communication interface is coupled with the processor, the processor is used for running programs or instructions to realize each process of the above-mentioned three-dimensional temperature field reconstruction method for an electrolytic cell based on multi-physical field coupling and achieve the same technical effects. To avoid repetition, details are not described herein.

[0095] It should be understood that the chip mentioned in the embodiments of the present application can also be referred to as a device-level chip, a device chip, a chip device, or a system-on-chip device, etc.

[0096] The above-mentioned embodiments are only a preferred scheme of the present application, and do not limit the present application in any form. Other variants and modifications can be made without exceeding the technical scheme recorded in the claims.

Claims

1. A method for reconstructing a three-dimensional temperature field of an electrolytic cell based on multi-physical field coupling, characterized in that, The application relates to a method for visualizing a three-dimensional temperature field of an electrolytic cell. The method comprises the following steps: a three-dimensional model of the electrolytic cell is layered along a Z-axis direction, a layered database is established according to a layered result, and a reference layer is set; an initial temperature field distribution of the reference layer is generated based on an anisotropic radial basis function interpolation algorithm; an electro-thermal-fluid coupling equation is established according to the initial temperature field of the reference layer, the layered database is sampled, the electro-thermal-fluid coupling equation is solved according to sampling data, and a temperature field distribution of the reference layer is obtained; the temperature field distribution of the reference layer is layered and mapped to a three-dimensional space along the Z-axis direction, the temperature field distribution of the reference layer is corrected in combination with interlayer temperature attenuation and transverse thermal diffusion effects, and a target temperature field distribution of the reference layer is obtained; 2. The multi-physical field coupling based reconstruction method of a three-dimensional temperature field of an electrolytic cell according to claim 1, characterized in that, the target temperature field distribution is rendered, and a visualized image of the three-dimensional temperature field of the electrolytic cell is obtained. The method comprises the following steps: the three-dimensional model of the electrolytic cell is discretized into multiple two-dimensional sections along the Z-axis direction; the multiple two-dimensional sections are adaptively layered according to curvature analysis and normal vector clustering algorithms in combination with a vertex curvature adaptive layering strategy; 3. The multi-physical field coupling based reconstruction method of a three-dimensional temperature field of an electrolytic cell according to claim 1, characterized in that, three-dimensional coordinate data of each layer is extracted, a layered database is established, and a reference layer is selected from the layers. The method comprises the following steps: a main diffusion axis in a flow passage section direction is obtained through principal component analysis, and an anisotropic distance measurement function is established; 4. The multi-physical field coupling based reconstruction method of a three-dimensional temperature field of an electrolytic cell according to claim 1, characterized in that, the initial temperature field distribution of the reference layer is generated by spatially interpolating each data in the layered database based on the anisotropic distance measurement function and in combination with a regularization parameter optimization. The method comprises the following steps: the reference layer is divided into four quadrants, temperature gradient averages in the quadrants are calculated, and weights of the quadrants are distributed according to the temperature gradient averages; sampling point numbers of the quadrants are distributed based on the corresponding weights, and additional sampling points are generated in combination with a boundary protection mechanism; 5. The multi-physical field coupling based reconstruction method of a three-dimensional temperature field of an electrolytic cell according to claim 1, characterized in that, the layered database is sampled according to the corresponding sampling point numbers and the generated additional sampling points, and sampling data are obtained. The method comprises the following steps: the electro-thermal-fluid coupling equation is discretized through a finite volume method, and an artificial diffusion term is introduced; 6. The multi-physical field coupling based reconstruction method of a three-dimensional temperature field of an electrolytic cell according to claim 1, characterized in that, soft constraint boundary conditions are set, the sampling data are taken as boundary inputs, the electro-thermal-fluid coupling equation after the discretization and the introduction of the artificial diffusion term is iteratively solved, and the temperature field distribution of the reference layer is obtained. The method comprises the following steps: temperature attenuation coefficients of the temperature field distribution of the reference layer along the Z-axis are calculated, and a spatial mapping relationship between non-reference layer vertices and the reference layer is established according to a fast nearest neighbor search algorithm; 7. The multi-physical field coupling based reconstruction method of a three-dimensional temperature field of an electrolytic cell according to claim 1, characterized in that, the temperature field distribution of the reference layer is corrected according to the temperature attenuation coefficients and a transverse thermal diffusion effect, and a target temperature layer distribution of the reference layer is obtained. The method comprises the following steps: The target temperature field distribution of the reference layer is rendered based on a multi-modal rendering engine to obtain a surface isotherm cloud chart and a three-dimensional volume rendering chart of the electrolytic cell.

8. A system for reconstructing a three-dimensional temperature field of an electrolytic cell based on multi-physical field coupling, configured to perform the method for reconstructing a three-dimensional temperature field of an electrolytic cell based on multi-physical field coupling according to any one of claims 1 to 7. The method comprises the following steps: An acquisition module is configured to acquire a three-dimensional model of the electrolytic cell, layer the three-dimensional model in the Z-axis direction, and establish a layering database and set a reference layer based on the layering result; A first processing module is configured to generate an initial temperature field distribution of the reference layer based on an anisotropic radial basis function interpolation algorithm; A second processing module is configured to sample the layering database to obtain corresponding sampling data; A third processing module is configured to input the sampling data into an electro-thermal-fluid coupling equation to solve the initial temperature field distribution of the reference layer and obtain the temperature field distribution of the reference layer; A fourth processing module is configured to layer map the temperature field distribution of the reference layer to a three-dimensional space along the Z-axis direction, correct the temperature field distribution of the reference layer in combination with interlayer temperature attenuation and lateral thermal diffusion effects, and obtain the target temperature field distribution of the reference layer; A rendering module is configured to render the target temperature field distribution of the reference layer to obtain a visualized image of the three-dimensional temperature field of the electrolytic cell.

9. An electronic device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, The processor executes the program to implement the method for reconstructing a three-dimensional temperature field of an electrolytic cell based on multi-physical field coupling according to any one of claims 1 to 7.

10. A non-transitory computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the method for reconstructing a three-dimensional temperature field of an electrolytic cell based on multi-physical field coupling according to any one of claims 1 to 7.

Images