SOEC pile variable-scale modeling simulation method and system based on data driving

Abstract

The invention discloses a data-driven SOEC electric pile variable-scale modeling simulation method and a data-driven SOEC electric pile variable-scale modeling simulation system. The method comprises the following operation steps of: establishing a three-dimensional multi-physical field coupling model of the SOEC single electrolytic tank; establishing an electrolytic tank agent model based on the SOEC single electrolytic tank three-dimensional multi-physics field coupling model; establishing a three-dimensional geometric model of the SOEC stack; defining multiple physical fields in the stack and setting an agent model, and establishing a three-dimensional multi-physical field coupling model of the SOEC stack; based on a finite element algorithm, SOEC electrolytic cell-pile variable-scale simulation is carried out, and the accuracy of the model is verified by combining pile performance experiment data; and the pile numerical simulation model and the intelligent optimization algorithm are combined to realize optimization of the pile operation parameters and structure. While the model calculation precision and accuracy are ensured, the model calculation amount can be effectively reduced, and the calculation efficiency is improved; on the premise that hardware computing power is not changed, modeling simulation of a larger-scale galvanic pile can be efficiently achieved.

Description

Technical Field

[0001] The present invention belongs to the field of solid oxide batteries, and relates to a data-driven SOEC stack scale-variable modeling simulation method and system; specifically, it relates to a data-driven solid oxide electrolytic cell stack scale-variable modeling simulation method and system. Background Art

[0002] Solid Oxide Electrolysis Cell (SOEC) is an all-solid-state energy conversion device that can convert electrical energy and thermal energy into chemical energy and store it in the product. It has the characteristics of fast reaction rate, high energy conversion efficiency, and environmental friendliness, showing good application prospects.

[0003] Compared with low-temperature electrochemical conversion technology using alkaline solution or proton exchange membrane as electrolyte, high-temperature electrochemical conversion technology using solid oxide as electrolyte has lower power consumption, faster electrode reaction rate, lower electrode polarization voltage, and less irreversible energy loss during electrolysis. Due to its higher temperature and reaction rate, SOEC does not need to use precious metal electrodes such as Pt, which greatly reduces the preparation cost. SOEC is a fully solid-state structure, which is easy to separate gas products and avoids the problems of evaporation, corrosion and electrolyte loss that may be caused by the use of liquid electrolytes.

[0004] SOEC can be regarded as the reverse process of solid oxide fuel cell (SOFC), which mainly consists of three parts: porous anode, dense electrolyte layer and porous anode. The battery unit generally also includes components such as cathode and anode flow channels, sealing layer, connector, metal frame, etc. In order to meet the gas production requirements of practical applications, multiple electrolytic cell units are usually assembled and stacked in series and parallel to form an electrolytic cell stack, and auxiliary components such as heating system, insulation device, air flow pipeline and gas distribution chamber are additionally configured. However, after assembling the monolithic electrolytic cell into a stack, the performance will be seriously reduced. This is because the stack structure is complex, and the stacking arrangement of the monolithic electrolytic cell will lead to inconsistency of component concentration, flow field, thermal field and electrochemical field in the stack, which will lead to thermal mismatch between components inside the stack and stress generated by temperature gradient. And as the scale of the stack increases, the inconsistency of the time / space distribution of state parameters in the stack and the difference in new energy between components are further enhanced, resulting in a decrease in the stability and durability of the stack.

[0005] At present, the problems that still exist in the research of solid oxide electrolytic cell stacks include: the study of the distribution characteristics of the internal state parameters of the stack, the clarification of the decisive operating conditions and structural parameters, the exploration of the scale-up rules of the stack, and the effective means to improve the performance of the stack. Limited by the long experimental research cycle and high cost, it is difficult to conduct a large number of geometric and operating parameters on the stack. In addition, since SOEC needs to be strictly sealed and operate under high temperature conditions, the physical field distribution inside the electrolytic cell is difficult to measure experimentally. The use of numerical simulation technology can simulate and analyze the physical and chemical processes inside the stack through full coupling simulation of multiple physical fields, obtain the details of the physical field distribution inside the SOEC stack with high geometric resolution, realize the prediction of the actual situation inside the stack and study the key factors affecting the performance of the stack.

[0006] The structure of the SOEC stack is complex. When conducting three-dimensional numerical simulation of multi-physical field coupling, the amount of calculation is large, the computing power is high, and the calculation time is long. Therefore, it is extremely difficult to perform three-dimensional multi-physical field modeling and simulation of large-scale SOEC stacks. In order to reduce the amount of calculation, common simplification methods include: (1) introducing a large number of ideal conditions and approximate assumptions, ignoring some structural information of the stack. However, the simplified simulation results often have a large error with the actual situation inside the stack, and subsequent structural optimization research on the SOEC stack cannot be carried out; (2) directly reducing the physical field or reducing the coupling relationship between the physical fields during the modeling process, resulting in ignoring the influence of the coupling relationship between the physical fields on the simulation results, and unable to accurately predict the distribution of the physical fields in the stack. Summary of the invention

[0007] In view of the above problems, the purpose of the present invention is to address the defects of the prior art and propose a data-driven variable-scale modeling and simulation method and system for SOEC battery stacks. By embedding the battery proxy model in the computational domain of a single electrolytic cell inside the battery stack, the partial differential solution of part of the process in the original computational domain is converted into algebraic equations. While ensuring the calculation accuracy of the model and the reliability of the results, the number of iteration steps is reduced, and the total number of degrees of freedom required for the solution is reduced, thereby effectively reducing the model solution memory and improving the computing efficiency. Under the condition of the same hardware computing power, modeling, simulation, performance analysis and optimization of larger-scale battery stacks can be achieved.

[0008] The technical solution of the present invention is: a data-driven SOEC stack variable-scale modeling and simulation method described in the present invention, and its operating steps are as follows:

[0009] Step (1): Establish a three-dimensional multi-physics field coupling model of the SOEC single cell and verify the accuracy of the model by combining experimental data;

[0010] Step (2): Based on the three-dimensional multi-physics field coupling model of the SOEC single battery, multiple simulations under different operating conditions are carried out for each sampling point of the sample set to obtain corresponding simulation results, solve the function mapping relationship representing the input-output, and construct a battery proxy model;

[0011] Step (3): Establishing a three-dimensional geometric model of the SOEC stack;

[0012] Step (4): for the three-dimensional geometric model of the SOEC stack, define the corresponding physical field and then build a built-in battery proxy model to construct a three-dimensional multi-physical field coupling model of the SOEC;

[0013] Step (5): For the three-dimensional multi-physics field coupling model of the SOEC stack, a finite element algorithm is used to perform a variable-scale simulation of the SOEC battery-stack, and the accuracy of the model is verified by combining experimental data;

[0014] Step (6): Combine the SOEC stack numerical simulation model with the intelligent optimization algorithm to optimize the stack operating parameters and stack structure.

[0015] Furthermore, in step (1), the process of establishing the three-dimensional multi-physics field model of the SOEC single cell specifically includes:

[0016] (1.1): Constructing geometric model: The geometric model of the SOEC single cell is constructed according to the three-dimensional structure size of the actual solid oxide electrolysis cell, including porous cathode, electrolyte, porous anode, gas flow channels for transporting reactants and products, and connectors for conducting and collecting current;

[0017] (1.2): Input material physical parameters: In the SOEC single cell multi-physics field coupling model, it is necessary to define the physical parameters of the materials of each component of the electrolytic cell and the physical parameters of the reaction gas on the electrode side, including: electrode effective conductivity, exchange current density, thermal conductivity, constant pressure heat capacity, specific heat, density, dynamic viscosity, gas diffusion coefficient; among them, the physical parameters of the reaction gas on the cathode and anode sides can be defined as temperature-dependent functions according to the specific components;

[0018] (1.3): Setting control equations and boundary conditions: There are multiple reactions in SOEC that interact with each other and are strongly coupled with the complex heat and mass transfer processes of multiple components. When establishing a multi-physics field coupling model, it is necessary to set the corresponding physical field control equations on each calculation domain, including electrochemical reaction equations, chemical reaction equations, mass conservation equations, momentum conservation equations, energy conservation equations, and charge conservation equations; when setting boundary conditions, the actual heat exchange between the fuel cell stack and the external environment and the influence of thermal radiation inside the fuel cell stack should be considered, and the coupling of multiple physical fields is realized by calling equation source terms and common variables;

[0019] (1.4): Grid independence analysis and model verification: The three-dimensional multi-physics field coupling model of SOEC single cell is solved based on the finite element algorithm. After grid independence verification, the accuracy and reliability of the model are verified by combining experimental data.

[0020] Furthermore, in step (2), the input variables of the proxy model include: the average surface temperature of the anode reaction zone, the average surface temperature of the cathode reaction zone, the average mole fraction of oxygen on the anode reaction zone, and the average mole fraction of reaction gas on the cathode reaction zone; the output variables of the proxy model are the electrolytic cell reaction current and the electrochemical reaction heat absorption of the electrolytic cell.

[0021] Furthermore, in step (3), the geometric modeling of the SOEC stack includes the electrolytic cell core group, the stack heating element, the insulation material, the gas supply and exhaust pipes, and the gas distribution chamber, so as to ensure that the geometric model is consistent with the actual stack components and structure, and avoid the inherent error between the simulation results and the actual stack performance caused by the ideal simplification of the stack geometry;

[0022] The electrolytic cell core group is a group of SOEC unit cells arranged at equal intervals, and the geometric structure of each unit cell is consistent with the geometric structure of the SOEC single cell in step (1); the stack heating element is used to provide the heat required for the operation of the stack, ensuring that the battery core group in the stack is always in a high-temperature working environment, and by introducing the heating element, the stack heating process and the actual temperature field spatial distribution in the stack are simulated, thereby avoiding the setting of isothermal or adiabatic boundary conditions and causing inaccurate results; the thermal insulation material wraps the battery core group, reduces the heat dissipation caused by convection heat transfer between the core area and the environment, and plays a role in heat preservation; the gas supply and exhaust pipeline includes a reaction gas inlet pipeline, a reaction gas / generated gas outlet pipeline, an air inlet pipeline, and an air outlet pipeline; the gas distribution chamber includes a reaction gas distribution chamber and an air distribution chamber, which together with the gas supply pipeline determine the flow configuration setting of the gas in the stack.

[0023] Furthermore, in step (4), the specific steps of establishing the three-dimensional multi-physics field coupling model of the SOEC stack include:

[0024] (4.1): The physical property parameters in the input parameters are defined according to the material composition of each unit component of the electrolytic cell core group, the specific components of the reaction gas on the cathode and anode sides, the heating element material, the insulation material composition, and the gas pipeline material; the physical property parameters include electrode effective conductivity, exchange current density, thermal conductivity, constant pressure heat capacity, specific heat rate, density, dynamic viscosity, gas diffusion coefficient, and emissivity;

[0025] (4.2): Setting physical field control equations and proxy models: setting corresponding momentum conservation equations, mass conservation equations, energy conservation equations and proxy model functions on each computational domain of the three-dimensional geometric model of the SOEC stack; the electrolytic cell proxy model is set on each unit cell domain in the stack, and the total number of independent variables that need to be solved in the stack is reduced by calling the proxy model output results, thereby reducing the overall number of degrees of freedom of the simulation model;

[0026] (4.3): Setting boundary conditions: Setting corresponding boundary conditions on different boundaries of the three-dimensional geometric model of the SOEC stack; the boundary conditions include flow inlet boundary conditions, pressure outlet boundary conditions, gas component inlet boundary conditions, and temperature boundary conditions;

[0027] (4.4): Mesh the three-dimensional geometric model of the SOEC stack and perform mesh independence analysis.

[0028] Furthermore, in step (5), the step of simulating the battery-stack variable-scale model for the SOEC stack multi-physics field coupling model includes:

[0029] (5.1): Based on the finite element algorithm, solve the partial differential equations of the conservation equations corresponding to each calculation domain of the SOEC stack model and the algebraic equations corresponding to each calculation domain of the electrolytic cell;

[0030] (5.2): Model verification is carried out in combination with the experimental results of the battery stack;

[0031] (5.1): Analyze the simulation results under different operating conditions and structural designs to explore the key operating and structural parameters that affect the overall performance of the fuel cell stack and the inconsistency of the internal state parameter distribution.

[0032] Furthermore, in step (6), the key operation and structural parameters that affect the overall performance of the fuel cell stack and the inconsistency of the distribution of internal state parameters obtained in step (5) are selected as decision variables, and the optimization problem is considered within the allowable design range and feasible operating conditions for the stability, safety, economy and efficiency of the long-term operation of the fuel cell stack;

[0033] The intelligent optimization algorithm gives the specific values ​​of the relevant decision variables and inputs them into the simulation model. The simulation model is solved to obtain the corresponding calculation results, which are then fed back to the optimization algorithm. The algorithm iterates and solves according to the search direction of the objective function. After multiple simulation-optimization processes, the optimal solution to the optimization problem is obtained.

[0034] Furthermore, a data-driven SOEC stack variable-scale modeling and simulation system includes the following modules:

[0035] The first model building module: Based on the geometric structure of the SOEC single cell, according to the preset material and reaction gas physical parameters, physical field control equations and boundary conditions, a three-dimensional multi-physics field coupling model of the SOEC single cell is constructed;

[0036] The first simulation solution module: Based on the finite element algorithm, after meshing the three-dimensional geometric model of the SOEC single cell, the multi-physical field coupling model of the SOEC single cell is iteratively solved to obtain the numerical solution of the model and provide data samples for the electrolytic cell proxy model;

[0037] Proxy model construction module: based on the mapping relationship between each sampling point of the sample set and the calculation results of the three-dimensional multi-physics field coupling model of the SOEC single cell, the proxy models of the electrolytic cell reaction current and the electrochemical reaction heat source are respectively constructed;

[0038] The second model building module: Based on the geometric structure of the SOEC stack, according to the preset material and reaction gas physical properties, physical field control equations and boundary conditions, the above-mentioned electrolytic cell reaction current and electrochemical reaction heat source proxy models are built into the calculation domain of each single electrolytic cell in the stack, so as to establish a three-dimensional multi-physics field coupling model of the SOEC stack;

[0039] The second simulation solution module: Based on the finite element algorithm, after meshing the three-dimensional geometric model of the SOEC stack, combined with the above-mentioned proxy model function, the three-dimensional multi-physics field coupling model of the SOEC stack is iteratively solved to obtain the numerical solution related to the model performance, and the key operation and structural parameters that affect the stack operation performance and the inconsistency of the internal state parameter distribution are obtained by comparative analysis of the key parameters;

[0040] Optimization module: Based on the improved SOEC stack multi-physics field coupling model, the intelligent optimization algorithm is called to optimize the stack operating parameters and stack configuration and structural parameters.

[0041] The beneficial effects of the present invention are as follows: 1. When modeling a SOEC unit cell, the present invention accurately restores the actual structure of the cell, takes into account the physical parameters of the mixed gas and the influence of the ambient temperature, takes into account the mutual coupling between different physical fields, compares and verifies the simulation results with the experimental results, and determines that the model prediction fits the actual situation; 2. Based on experimental design and proxy model technology, the present invention constructs a data sample set through multiple numerical simulations of SOEC single cells, and establishes a SOEC unit cell proxy model based on the input-output mapping relationship, so as to quickly and easily obtain the gas production performance and heat absorption of the electrolytic cell under different operating conditions, thereby reducing the calculation cost and improving the calculation efficiency, and the corresponding calculation is completed in seconds; 3. The present invention accurately models the battery stack according to the actual structure, and the geometric model includes the actual structure of the auxiliary components, taking into account the influence of air convection heat transfer and radiation heat transfer in a high temperature environment on the temperature field in actual conditions, rather than simplifying it to isothermal or adiabatic boundary conditions. , making the simulation model construction closer to reality and the prediction results more in line with the actual performance of the battery stack; 4. The present invention builds a proxy model on each unit electrolytic cell in the SOEC battery stack model, and replaces the partial differential solution of charge conservation, momentum conservation, energy conservation, electrochemical reaction and other processes in the original electrolytic cell calculation domain with algebraic equations. While ensuring the accuracy and reliability of the model calculation, it reduces the iterative convergence steps in the calculation and reduces the total number of degrees of freedom required for the solution, thereby effectively reducing the model solution memory and improving the calculation efficiency. Under the same hardware computing power conditions, larger-scale battery stacks can be modeled, simulated, analyzed and optimized; 5. Based on an intelligent optimization algorithm, the present invention conducts joint simulation optimization of the above-mentioned accurate and efficient multi-physical field coupled three-dimensional SOEC battery stack model through program code to study the problem of battery stack performance optimization; wherein, the optimization algorithm and the simulation model call each other, thereby realizing the optimization of the target SOEC battery stack operating conditions and structural design. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 is an operation flow chart provided by an embodiment of the present invention;

[0043] Figure 2 is a specific flow chart of step 1 in the method provided in an embodiment of the present invention;

[0044] Figure 3 is a specific flow chart of step 2 in the method provided in an embodiment of the present invention;

[0045] Figure 4 is a specific flow chart of step 4 in the method provided in an embodiment of the present invention;

[0046] Figure 5 It is a structural schematic diagram of a data-driven solid oxide electrolytic cell stack variable-scale modeling and simulation system provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0047] The specific technical scheme of the present invention is further described in detail below with reference to specific examples.

[0048] Example 1

[0049] This example provides a data-driven SOEC stack variable-scale modeling and simulation method. The specific process is as follows: Figure 1 As shown, the following steps are included:

[0050] Step (1): Establish a three-dimensional multi-physics field coupling model of a SOEC single cell;

[0051] Step (2): Based on the three-dimensional multi-physics field coupling model of the SOEC single cell, select sampling points of the sample set, obtain the simulation results of the selected sampling points under different operating conditions, solve the mapping relationship representing the input-output function, and realize the establishment of the electrolytic cell proxy model;

[0052] Step (3): Based on the actual construction of the fuel cell stack, a three-dimensional geometric model of the SOEC fuel cell stack including auxiliary components is established;

[0053] Step (4): define the corresponding physical field for the SOEC stack geometric model, and build in the electrolytic cell proxy model to establish a three-dimensional multi-physical field coupling model of the SOEC stack;

[0054] Step (5): Based on the finite element algorithm, a SOEC single electrolytic cell-stack variable scale simulation is carried out for the above-mentioned SOEC stack three-dimensional multi-physics field coupling model, and the simulation results are compared and verified with the stack performance experimental results to ensure the accuracy of the model;

[0055] Step (6): Combining the SOEC stack numerical simulation model with the intelligent optimization algorithm, an optimization scheme for the stack operating parameters and stack structure is proposed.

[0056] Furthermore, the step (1) is specifically implemented according to the following steps:

[0057] (1.1) Constructing geometric model: The geometric model of the SOEC single cell is constructed according to the three-dimensional structure dimensions of the actual solid oxide electrolysis cell, including a porous cathode, electrolyte, porous anode, gas flow channels for transporting reactants, and connectors for conducting and collecting current;

[0058] The porous cathode comprises a cathode functional layer and a cathode support layer, wherein the cathode functional layer is in direct contact with the electrolyte layer, plays a role in catalyzing the reaction gas and providing a reaction site, and the support layer plays a role in supporting and transmitting the gas; the porous anode comprises an anode functional layer and an anode support layer; the gas flow channel comprises a reaction gas flow channel on the cathode side and an air flow channel on the anode side; the connector is required to have strong conductivity, usually made of highly conductive metal or alloy material, for conducting and collecting reaction current and connecting to an external circuit;

[0059] (1.2) Physical property parameters of input materials: material physical property parameters are defined according to the specific material composition of each component of the solid oxide electrolytic cell, and mixed gas physical property parameters are defined according to the specific components and temperature and pressure conditions of the reaction gas in the cathode, anode and gas flow pipeline; the physical property parameters include electrode effective conductivity, electrolyte conductivity, electrochemical reaction exchange current density, thermal conductivity, constant pressure heat capacity, specific heat, density, dynamic viscosity, gas diffusion coefficient;

[0060] Among them, the physical properties of porous electrodes need to take into account the combined effects of the solid phase of the composite material and the gas phase in the pores; the effects of different components on the physical properties of the mixed reaction gas need to be considered, and the temperature dependence function of the physical properties of the materials of each component of the electrolytic cell and the mixed gas should be determined through physical property experimental data;

[0061] (1.3) Setting control equations: For the three-dimensional geometric model of the SOEC single cell established above, corresponding conservation equations are set on each computational domain of the geometric model according to the electrochemical reactions, chemical reactions, momentum transfer, mass transfer, energy transfer and charge transfer occurring during the operation of the electrolytic cell;

[0062] The reaction is described in the form of equations, which are respectively set on the cathode functional layer and the anode functional layer, taking into account the electrochemical reaction and the catalytic reforming reaction; the momentum conservation equation takes into account the flow of the mixed gas in the porous electrode and the gas flow channel; the mass conservation equation takes into account the generation and consumption of substances caused by electrochemical reactions and chemical reactions in the cathode and anode functional layers, and the gas diffusion mode takes into account the combination of molecular diffusion and Crusen diffusion processes; the energy conservation equation takes into account the processes of heat conduction, heat convection and heat radiation in the entire battery calculation domain, as well as the electrochemical reaction endothermicity and the ohmic heat of the electrolytic cell; the charge conservation equation is set on the cathode, anode, electrolyte and connector of the electrolytic cell, taking into account the conservation of ions and electrons on the electrolyte and electrodes, and the contact resistance of the connector and the electrode contact surface;

[0063] By calling the source terms and common variables in the control equations, the mutual coupling of the above different processes is achieved;

[0064] (1.4) Setting boundary conditions: For the above-mentioned three-dimensional geometric model, corresponding boundary conditions are set according to the setting results of the physical field in the computational domain. The boundary conditions include the flow inlet boundary conditions of the mixed gas, the gas component inlet boundary conditions, the pressure outlet boundary conditions, the potential boundary conditions, and the temperature boundary conditions;

[0065] (1.5) Grid division: Grid the above three-dimensional geometric model, perform partial differential equation calculation on the above multi-physics field coupling model based on the finite element algorithm, iteratively solve until convergence, and obtain the numerical solution corresponding to the model; conduct grid independence analysis, compare the fluctuation of calculation results under different grid numbers, and select the appropriate grid number;

[0066] (1.6) Verify the accuracy of SOEC model simulation by combining experimental data; conduct performance experiments on SOEC single cells, measure the volt-ampere characteristic data of the electrolytic cell under standard operating conditions, and verify the accuracy of the model calculation results.

[0067] like Figure 3 As shown, step (2) is specifically described as:

[0068] (2.1) The proxy model input variables include: the average surface temperature of the anode of the electrolytic cell, the average surface temperature of the cathode, the average mole fraction of oxygen on the surface of the anode functional layer, and the average mole fraction of the reaction gas on the cathode functional layer; the proxy model output variables include: the reaction current of the electrolytic cell and the heat absorbed by the electrochemical reaction of the electrolytic cell;

[0069] (2.2) Select sampling points: determine the allowed value range of each variable, construct sample set and test set sampling points according to the above value range, and perform random sampling in the design space based on the experimental design method. The number of sampling points is 20N and 10N respectively, where N is the quantity variable. The experimental design method can select any random sampling method;

[0070] (2.3) Define the input and output parameters in the SOEC three-dimensional multi-physics coupling model;

[0071] The input parameters include the working voltage of the electrolytic cell, the average surface temperature of the electrode functional layer, the average mole fraction of the reaction gas in the electrode functional layer, and the reaction gas inlet flow rate. The output parameters are the current and electrochemical reaction heat absorption of the SOEC single cell model under the corresponding operating conditions; the input-output variables of the simulation model are consistent with the proxy model;

[0072] (2.4) Based on the above sample set and each sampling point of the test set, the S23 simulation calculation is repeated multiple times, and the electrolytic cell simulation calculation under different working conditions is carried out to obtain the corresponding simulation results of each sampling point, and establish a mapping relationship representing the input-output;

[0073] (2.5) Based on the mapping relationship between the sampling points of the above sample set and the simulation calculation results, the electrolytic cell reaction current and electrochemical reaction heat source proxy models are established respectively, and the functional relationship is listed as follows:

[0074] I cell =f Icell (V cell , T av , x av )

[0075] Q cell =f Qelec (V cell , T av , x av )

[0076] In the formula, I cell is the electrolytic cell reaction current, Q elec is the heat source of electrochemical reaction in the electrolytic cell, V cell is the operating voltage of the electrolytic cell, T av is the average surface temperature of the electrode functional layer, x av is the average mole fraction of the gas component in the electrode functional layer, and f is the proxy model function relationship of the mapping model input-output;

[0077] Based on the above process, verify whether the accuracy of the proxy model meets the error requirements. If not, increase the number of sample set collection points, perform simulation calculations again, and reconstruct the proxy model function relationship based on the calculation results; the proxy model function form can be any of a polynomial form, a radial basis function form, a response surface form or other forms.

[0078] In step (3), the geometric modeling of the SOEC stack includes the electrolytic cell core group, the stack heating element, the insulation material, the gas supply and exhaust pipes, and the gas distribution chamber, so as to ensure that the geometric model is consistent with the actual stack components and structure, and avoid the inherent error between the simulation results and the actual stack performance caused by the ideal simplification of the stack geometry;

[0079] The electrolytic cell core group is a group of SOEC unit cells arranged at equal intervals, and the geometric structure of each unit cell is consistent with the geometric structure of the SOEC single cell in step (1); the stack heating element is used to provide the heat required for the operation of the stack, ensuring that the battery core group in the stack is always in a high-temperature working environment, and by introducing the heating element, the stack heating process and the actual temperature field spatial distribution in the stack are simulated, thereby avoiding the setting of isothermal or adiabatic boundary conditions and causing inaccurate results; the thermal insulation material wraps the battery core group, reduces the heat dissipation caused by convection heat transfer between the core area and the environment, and plays a role in heat preservation; the gas supply and exhaust pipeline includes a reaction gas inlet pipeline, a reaction gas / generated gas outlet pipeline, an air inlet pipeline, and an air outlet pipeline; the gas distribution chamber includes a reaction gas distribution chamber and an air distribution chamber, which together with the gas supply pipeline determine the flow configuration setting of the gas in the stack.

[0080] like Figure 4 As shown, step (4) is specifically described as:

[0081] (4.1) Define input parameters: The physical property parameters in the input parameters are defined according to the material composition of each unit component of the electrolytic cell core group, the specific components of the cathode and anode side reaction gases, the heating element material, the insulation material composition, and the gas pipeline material; the physical property parameters include electrode effective conductivity, exchange current density, thermal conductivity, constant pressure heat capacity, specific heat rate, density, dynamic viscosity, gas diffusion coefficient, and emissivity;

[0082] Furthermore, the physical properties of porous electrodes need to take into account the combined effects of the solid phase of the composite material and the gas phase in the pores; the effects of different components on the physical properties of the mixed reaction gas need to be considered, and the temperature dependence function of the physical properties of the materials of each component of the electrolytic cell and the mixed gas should be determined through physical property experimental data;

[0083] (4.2) Setting physical field control equations and proxy models: setting corresponding momentum conservation equations, mass conservation equations, energy conservation equations and proxy model functions on each calculation domain of the three-dimensional geometric model of the SOEC stack;

[0084] The momentum conservation equation takes into account the flow process of the reaction gas into the pipeline, the gas distribution chamber, the internal flow channel of the electrolytic cell group and the gas in the porous electrode;

[0085] The mass conservation equation takes into account the molecular diffusion and Knudsen diffusion processes in the reaction gas inflow pipeline, gas distribution chamber, internal flow channel of the electrolytic cell group and porous electrodes;

[0086] The energy conservation equation takes into account the heat conduction and convection processes in the entire stack calculation domain. Since the SOEC stack operates at high temperature, the radiation heat exchange between the surface of the heating element and the surface and internal wall of the electrolytic cell must also be considered;

[0087] The electrolytic cell proxy model includes an electrolytic cell reaction current and an electrochemical reaction heat source proxy model. By solving the proxy model set in each electrolytic cell domain in the stack instead of solving the electrochemical reaction equation, the secondary current equation and the partial differential equation of electrochemical heat absorption in the electrolytic cell calculation domain, the total number of independent variables that need to be solved and the number of degrees of freedom of the model as a whole can be reduced, thereby improving the calculation efficiency.

[0088] Furthermore, the proxy model input variables include the average surface temperature of the cathode functional layer of each electrolytic cell, the average surface temperature of the anode functional layer, the average mole fraction of the reaction gas on the cathode functional layer, and the average mole fraction of oxygen on the anode functional layer. The output variables are the reaction current and electrochemical heat generation of each electrolytic cell, and are coupled to the mass transfer field and the temperature field as source terms.

[0089] Furthermore, the consumption of reaction gas on the cathode and oxygen and the generation of product gas can be solved by the output current result of the above proxy model, and the gas source term can be substituted into the mass conservation equation. Taking the SOEC stack for hydrogen production by electrolysis of water as an example, the source term Smass of oxygen generation on the anode side and hydrogen generation on the cathode side can be calculated by the following formula:

[0090]

[0091] In the formula, M is the molar mass, F is the Faraday constant, and S is TPB，ca is the effective active area of ​​the cathode, S TPB，an is the effective active area of ​​the anode;

[0092] (4.3) Set the boundary conditions of the model: For the three-dimensional multi-physics field coupling model of the SOEC stack mentioned above, set the corresponding boundary conditions on different boundaries of the three-dimensional geometric model, including gas flow inlet boundary conditions, gas component inlet boundary conditions, pressure outlet boundary conditions, and temperature boundary conditions. In addition, the secondary current of the stack can be directly solved algebraically through the functional relationship of the proxy model, so the SOEC stack simulation model with built-in proxy model does not need to set additional potential boundary conditions;

[0093] (4.4) The three-dimensional geometric model of the above-mentioned SOEC stack is meshed and a mesh independence analysis is performed.

[0094] Furthermore, step (5) is specifically described as follows:

[0095] (5.1) Based on the finite element algorithm, solve the partial differential equations of the algebraic equations corresponding to the calculation domain of each electrolytic cell and the physical field control equations corresponding to each calculation domain of the stack model;

[0096] (5.2) Combined with the experimental data of SOEC stack performance, the model is verified to ensure its accuracy;

[0097] (5.3) Through multiple solutions, the simulation results under different operating conditions and allowable structural designs are obtained, and key parameter analysis is performed to explore the key operating and structural parameters that affect the overall performance of the fuel cell stack and the inconsistency of the internal state parameter distribution.

[0098] Furthermore, in step (6), the key operation and structural parameters that affect the overall performance of the fuel cell stack and the inconsistency of the distribution of internal state parameters obtained in step (5) are selected as decision variables, and the optimization problem is considered within the allowable design range and feasible operating conditions for the stability, safety, economy and efficiency of the long-term operation of the fuel cell stack;

[0099] Furthermore, the specific values ​​of the relevant decision variables are given by the intelligent optimization algorithm, input into the simulation model for solution, and then the obtained calculation results are fed back to the optimization algorithm, and the solution is iterated according to the search direction of the objective function. After multiple simulation-optimization processes, the optimal solution to the optimization problem is obtained.

[0100] Example 2

[0101] This embodiment provides a data-driven SOEC stack variable-scale modeling and simulation system, such as Figure 5 As shown, it contains the following modules:

[0102] The first model building module: Based on the three-dimensional geometric structure of the SOEC single cell, a multi-physical field coupling model of the SOEC single cell is established according to the set material properties, reactant and product properties, physical field control equations and corresponding boundary conditions.

[0103] The first simulation solution model: mesh the multi-physical field coupling model of the SOEC single cell, iteratively solve the above model based on the finite element algorithm, obtain the numerical solution of the model, and provide data samples for constructing the electrolytic cell proxy model.

[0104] Proxy model construction module: Based on the mapping relationship between each sampling point in the sample set and the solution results of the SOEC single cell multi-physics field coupling model, the electrolytic cell reaction current and electrochemical reaction heat source proxy models are constructed respectively.

[0105] The second model building module: Based on the actual geometric structure of the SOEC battery stack, according to the set material properties, reactant and product properties, physical field control equations and corresponding boundary conditions, the above-mentioned battery proxy model is set on the corresponding electrolytic cell calculation domain to establish a SOEC single battery multi-physical field coupling model.

[0106] The second model solving module is to mesh the multi-physics coupling model of the SOEC stack, iteratively solve the multi-physics coupling model based on the finite element algorithm and combined with the proxy model function to obtain the numerical solution of the model, and obtain the key operation and structural parameters that affect the stack performance and the inconsistent distribution of internal parameters through key parameter analysis.

[0107] Optimization module: Based on the improved three-dimensional multi-physics field coupling model of the SOEC stack, the intelligent optimization algorithm is called to optimize the stack operating parameters and stack structure.

Claims

1. A data-driven SOEC stack variable-scale modeling and simulation method, characterized in that: The operation steps are as follows: Step (1): Establish a three-dimensional multi-physics field coupling model of the SOEC single cell and verify the accuracy of the model by combining experimental data; Step (2): Based on the three-dimensional multi-physics field coupling model of the SOEC single battery, multiple simulations under different operating conditions are carried out for each sampling point of the sample set to obtain corresponding simulation results, solve the function mapping relationship representing the input-output, and construct a battery proxy model; Step (3): Establishing a three-dimensional geometric model of the SOEC stack; Step (4): for the constructed three-dimensional geometric model of the SOEC stack, define the corresponding physical field and the built-in battery proxy model to construct a three-dimensional multi-physical field coupling model of the SOEC; Step (5): For the three-dimensional multi-physics field coupling model of the SOEC stack, a finite element algorithm is used to perform a variable-scale simulation of the SOEC battery-stack, and the accuracy of the model is verified by combining experimental data; Step (6): Combine the SOEC stack numerical simulation model with the intelligent optimization algorithm to optimize the stack operating parameters and stack structure.

2. The data-driven variable-scale modeling and simulation method for SOEC stack according to claim 1, characterized in that: In step (1), the specific method of constructing the three-dimensional multi-physics field coupling model of the SOEC single cell includes: The geometric model of a single SOEC cell is constructed based on the three-dimensional structural dimensions of an actual solid oxide electrolysis cell, including a porous cathode, electrolyte, porous anode, gas flow channels for transporting reactants and products, and connectors for conducting and collecting current. The physical properties of the materials of each component of the electrolytic cell and the physical properties of the reaction gas on the electrode side need to be defined in the multi-physics coupling model of the SOEC single cell, including: effective conductivity of the electrode, exchange current density, thermal conductivity, constant pressure heat capacity, specific heat, density, dynamic viscosity, and gas diffusion coefficient; Among them, the physical properties of the reaction gases on the cathode and anode sides can be defined as temperature-dependent functions according to the specific components; There are multiple reactions in SOEC that interact with each other and are strongly coupled with the complex heat and mass transfer processes of multiple components. When establishing a multi-physics field coupling model, it is necessary to set the corresponding physical field control equations on each calculation domain, including the electrochemical reaction equation, chemical reaction equation, mass conservation equation, momentum conservation equation, energy conservation equation, and charge conservation equation. When setting boundary conditions, the actual heat exchange between the fuel cell stack and the external environment and the influence of thermal radiation inside the fuel cell stack are considered, and the coupling of multiple physical fields is achieved by calling equation source terms and shared variables; The three-dimensional multi-physics field coupling model of SOEC single cell is solved based on the finite element algorithm. After grid independence verification, the accuracy and reliability of the model are verified by combining experimental data.

3. The data-driven variable-scale modeling and simulation method for SOEC stack according to claim 1, characterized in that: In step (2), the input variables of the constructed battery proxy model include: the average surface temperature of the anode reaction zone, the average surface temperature of the cathode reaction zone, the average mole fraction of oxygen on the anode reaction zone, and the average mole fraction of reaction gas on the cathode reaction zone; the output variables of the proxy model are the electrolytic cell reaction current and the electrochemical reaction endothermic amount of the electrolytic cell.

4. The data-driven variable-scale modeling and simulation method for SOEC stack according to claim 1 is characterized in that: In step (3), the three-dimensional geometric modeling of the SOEC stack established includes the electrolytic cell core group, the stack heating elements, the insulation material, the gas supply and exhaust pipes, and the gas distribution chamber, to ensure that the geometric model is consistent with the actual stack components and structure, and to avoid the inherent error between the simulation results and the actual stack performance caused by the ideal simplification of the stack geometry.

5. The data-driven variable-scale modeling and simulation method for SOEC stack according to claim 4 is characterized in that: in, The electrolytic cell core group is a group of equidistantly arranged SOEC unit cells, and the geometric structure of each unit cell is consistent with the geometric structure of the SOEC single cell in step (1); The stack heating element is used to provide the heat required for the operation of the stack, ensuring that the battery core group in the stack is always in a high-temperature working environment. By introducing the heating element, the stack heating process and the actual temperature field spatial distribution in the stack are simulated to avoid inaccurate results caused by setting isothermal or adiabatic boundary conditions; The thermal insulation material wraps the battery core pack to reduce heat dissipation caused by convection heat exchange between the core area and the environment; The gas supply and exhaust pipeline includes a reaction gas inlet pipeline, a reaction gas / product gas outlet pipeline, an air inlet pipeline, and an air outlet pipeline; The gas distribution chamber includes a reaction gas distribution chamber and an air distribution chamber, which together with the gas supply pipeline determine the flow configuration of the gas in the stack.

6. The data-driven variable-scale modeling and simulation method for SOEC stack according to claim 1, characterized in that: In step (4), the specific steps of establishing the three-dimensional multi-physics field coupling model of the SOEC stack are as follows: (4.1): The physical property parameters in the input parameters are defined according to the material composition of each unit component of the electrolytic cell core group, the specific components of the reaction gas on the cathode and anode sides, the heating element material, the insulation material composition, and the gas pipeline material; the physical property parameters include electrode effective conductivity, exchange current density, thermal conductivity, constant pressure heat capacity, specific heat rate, density, dynamic viscosity, gas diffusion coefficient, and emissivity; (4.2): Setting physical field control equations and proxy models: setting corresponding momentum conservation equations, mass conservation equations, energy conservation equations and proxy model functions on each computational domain of the three-dimensional geometric model of the SOEC stack; the electrolytic cell proxy model is set on each unit cell domain in the stack, and the total number of independent variables that need to be solved in the stack is reduced by calling the proxy model output results, thereby reducing the overall number of degrees of freedom of the simulation model; (4.3): Setting boundary conditions: Setting corresponding boundary conditions on different boundaries of the three-dimensional geometric model of the SOEC stack; The boundary conditions include flow inlet boundary conditions, pressure outlet boundary conditions, gas component inlet boundary conditions, and temperature boundary conditions; (4.4): Mesh the three-dimensional geometric model of the SOEC stack and perform mesh independence analysis.

7. The data-driven variable-scale modeling and simulation method for SOEC stack according to claim 1, characterized in that: In step (5), the step of constructing a SOEC stack multi-physics field coupling model battery-stack variable scale model simulation includes: (5.1): Based on the finite element algorithm, solve the partial differential equations of the conservation equations corresponding to each calculation domain of the SOEC stack model and the algebraic equations corresponding to each calculation domain of the electrolytic cell; (5.2): Model verification is carried out in combination with the experimental results of the battery stack; (5.3): Analyze the simulation results under different operating conditions and structural designs to explore the key operating and structural parameters that affect the overall performance of the fuel cell stack and the inconsistency of the internal state parameter distribution.

8. The data-driven variable-scale modeling and simulation method for SOEC stack according to claim 1, characterized in that: In step (6), the method of optimizing the stack operating parameters and stack structure is specifically: The key operation and structural parameters that affect the overall performance of the fuel cell stack and the inconsistency of the distribution of internal state parameters obtained in step (5) are selected as decision variables, and the optimization problem is considered within the allowable design range and feasible operating conditions for the stability, safety, economy and efficiency of the long-term operation of the fuel cell stack; The intelligent optimization algorithm gives the specific values ​​of the relevant decision variables and inputs them into the simulation model. The simulation model is solved to obtain the corresponding calculation results, which are then fed back to the optimization algorithm. The algorithm iterates and solves according to the search direction of the objective function. After multiple simulation-optimization processes, the optimal solution to the optimization problem is obtained.

9. A data-driven SOEC stack variable-scale modeling and simulation system, characterized in that: It includes a first model building module, a first simulation solution module, an agent model building module, a second model building module, a second simulation solution module and an optimization module.

10. A data-driven SOEC stack variable-scale modeling and simulation system according to claim 1, characterized in that: Contains the following modules: The first model building module: based on the geometric structure of the SOEC single cell, according to the preset material and reaction gas physical properties, physical field control equations and boundary conditions, builds a three-dimensional multi-physical field coupling model of the SOEC single cell; The first simulation solution module: Based on the finite element algorithm, after meshing the three-dimensional geometric model of the SOEC single cell, the multi-physical field coupling model of the SOEC single cell is iteratively solved to obtain the numerical solution of the model and provide data samples for the electrolytic cell proxy model; Proxy model construction module: Based on the mapping relationship between each sampling point in the sample set and the calculation results of the three-dimensional multi-physics field coupling model of the SOEC single cell, the proxy models of the electrolytic cell reaction current and the electrochemical reaction heat source are constructed respectively; The second model building module: Based on the geometric structure of the SOEC stack, according to the preset material and reaction gas physical properties, physical field control equations and boundary conditions, the above-mentioned electrolytic cell reaction current and electrochemical reaction heat source proxy models are built into the calculation domain of each single electrolytic cell in the stack, so as to establish a three-dimensional multi-physics field coupling model of the SOEC stack; The second simulation solution module: Based on the finite element algorithm, after meshing the three-dimensional geometric model of the SOEC stack, combined with the above-mentioned proxy model function, the three-dimensional multi-physics field coupling model of the SOEC stack is iteratively solved to obtain the numerical solution related to the model performance, and the key operation and structural parameters that affect the stack operation performance and the inconsistency of the internal state parameter distribution are obtained by comparative analysis of the key parameters; Optimization module: Based on the improved SOEC stack multi-physics field coupling model, the intelligent optimization algorithm is called to optimize the stack operating parameters and stack configuration and structural parameters.

Images