A Multi-Objective Optimization Method for Thermal Insulation Topology Based on Iterative Correction of Fluid-Structure Coupling Thermal Networks

Abstract

This invention discloses a multi-objective optimization method for thermal insulation topology based on iterative correction of a fluid-structure interaction thermal network, belonging to the field of thermal control and regulation technology. The method includes: constructing an initial fluid-structure interaction thermal network model; performing reverse iterative correction of the convective heat transfer coefficient based on a reference temperature field until the model's predicted temperature field and the reference temperature field reach consistency under a convergence criterion; introducing lumped heat capacity elements at the nodes of the corrected thermal network model to establish a transient RC thermal network model; embedding the corrected thermal network model into a multi-objective optimization framework, using thermal insulation performance and structural quality as optimization objectives, and employing a multi-objective evolutionary algorithm to iteratively optimize the thermal insulation structure topology to obtain the Pareto optimal solution set; and using a fuzzy compromise decision criterion to select the optimal compromise solution. This invention effectively improves the computational accuracy of the thermal network method while maintaining its high efficiency, achieving performance-driven optimization of the thermal insulation structure.

Description

Technical Field

[0001] This invention relates to the field of thermal control and regulation technology, and in particular to a multi-objective optimization method for thermal insulation topology based on iterative correction of fluid-structure interaction thermal networks. Background Technology

[0002] In thermal protection and thermal control systems for aerospace, energy and power, and electronic equipment thermal management, the performance of the insulation structure directly affects the system's reliability and service life. With the increasing demands for lightweight, high-temperature resistance, and thermal protection in high-end equipment, how to efficiently and accurately predict and control the thermal state of the insulation system has become a key technical challenge in thermal control system design.

[0003] Currently, the main methods for monitoring and controlling the thermal state of insulation systems include analytical calculation methods, finite volume methods, and lumped parameter thermal network methods.

[0004] Analytical calculation methods obtain the temperature field distribution by solving the exact analytical solution of the heat conduction differential equation. However, this method requires extreme simplification of complex engineering problems. For practical thermal insulation systems with complex geometries, nonlinear boundary conditions, and temperature-dependent material properties, analytical methods are difficult to apply and cannot provide real-time and accurate state feedback for the control system.

[0005] The finite volume method can obtain a high-resolution temperature field distribution by spatially discretizing the computational domain. However, this method faces significant bottlenecks in real-time performance and computational efficiency when applied to practical thermal control systems: solving complex physical fields at high resolution requires a large number of iterative calculations, consuming huge computational resources and making it difficult to meet the timeliness requirements of real-time state monitoring and control decision-making in thermal control systems.

[0006] The lumped-parameter thermal network method, based on the thermoelectric analogy principle, constructs a resistance-capacitance network model for solution. It boasts significant advantages such as flexible modeling and high computational efficiency, making it suitable as a state prediction model for control systems. However, its core drawback in thermal control applications lies in its insufficient accuracy, particularly in convective heat transfer. The calculation of convective thermal resistance heavily relies on the selection of empirical formulas and coefficients. When actual operating conditions deviate from the applicability of these formulas, the thermal resistance estimation becomes significantly inaccurate, leading to deviations in the control system's temperature prediction from the true value, resulting in control overshoot or regulation failure. Furthermore, when dealing with transient heat conduction problems, the traditional thermal network method lacks a systematic and precise control mechanism for integrating heat capacity parameters.

[0007] In summary, existing state prediction methods in thermal control systems struggle to achieve a good balance between computational accuracy and efficiency. There is an urgent need to develop a thermal state control method for insulation systems that can maintain the high-efficiency computational advantages of the thermal network method while systematically improving prediction accuracy and achieving effective control and regulation. Summary of the Invention

[0008] The purpose of this invention is to provide a multi-objective optimization method for thermal insulation topology based on iterative correction of fluid-structure interaction thermal networks. By introducing dynamically corrected convective heat transfer coefficients and integrated heat capacity parameters, the method achieves accurate calibration of the thermal network model and embeds the corrected high-precision thermal network model into a multi-objective optimization framework. This solves the technical problems of insufficient accuracy caused by the reliance on empirical formulas in the calculation of convective thermal resistance in existing thermal network methods, as well as the lack of a systematic accuracy calibration and control mechanism.

[0009] To achieve the above objectives, this invention provides a multi-objective optimization method for thermal insulation topology based on iterative correction of fluid-structure interaction thermal networks, comprising the following steps: S1: Based on the physical configuration and material properties of the target thermal insulation structure, an initial fluid-structure coupled thermal network model is constructed using the lumped parameter method. The initial values ​​of the convective heat transfer coefficients at each solid-fluid interface in the thermal network model are estimated based on empirical formulas. S2: Using a high-fidelity reference temperature field as a benchmark, calculate the deviation between the temperature field predicted by the thermal network model and the reference temperature field. Based on the deviation, perform reverse iterative correction on the convective heat transfer coefficient until the temperature field predicted by the thermal network model and the reference temperature field reach consistency under a preset convergence criterion, and obtain the corrected thermal network model. S3: At each node of the modified thermal network model obtained in step S2, lumped heat capacity elements are introduced. The heat capacity value of each node is calculated based on the specific heat capacity, density and corresponding control volume of the material, and a transient RC thermal network model including thermal resistance elements and heat capacity elements is established. S4: The modified thermal network model obtained in step S2 is embedded into a multi-objective optimization framework as a solver. The thermal insulation performance index and structural quality index are used as optimization objectives, and the material density distribution in the design domain is used as the design variable. The multi-objective intelligent optimization algorithm is used to iteratively optimize the topology of the thermal insulation structure to obtain the Pareto optimal solution set. S5: From the Pareto optimal solution set obtained in step S4, select the optimal compromise solution according to the preset decision criteria, and output it as the final thermal insulation structure topology optimization scheme.

[0010] Preferably, in step S1, the initial fluid-structure interaction thermal network model is constructed as follows: the insulation structure is divided into multiple control volume units along the thickness direction, the average temperature of each control volume unit is characterized by the node temperature, adjacent nodes are connected by conductive thermal resistance, and the solid-fluid interface is connected by convective thermal resistance. The expression for the convective thermal resistance is: ; in, For convective thermal resistance, The convective heat transfer coefficient is... This represents the convective heat transfer area.

[0011] Preferably, in step S2, the reverse iterative correction of the convective heat transfer coefficient is as follows: in the first... In the next iteration, the convective heat transfer coefficient of each convective heat transfer node is corrected according to the following formula: ; in, Indicates the first One convection heat transfer node For the first In the nth iteration The convective heat transfer coefficient of each node, For the first In the nth iteration The convective heat transfer coefficient of each node, For the first Reference temperature values ​​at each node For the first In the nth iteration The model predicts the temperature value at each node. As a relaxation factor, This is the regularization parameter.

[0012] Preferably, in step S2, the preset convergence criterion is: the maximum value of the relative change in temperature of each node in two adjacent iterations is less than or equal to a first threshold. Meanwhile, the mean square value of the relative change in temperature at each node between two adjacent iterations is less than or equal to the second threshold. ,Right now: ; and ; in, This represents the total number of temperature monitoring nodes. For the first In the nth iteration The model predicts the temperature value at each node. The threshold for relative change in local temperature. This represents the threshold for relative change in mean square temperature.

[0013] Preferably, in step S2, the reference temperature field is obtained by any of the following methods: by calculating the target insulation structure after fine meshing using the finite volume method, or by numerical simulation of the target insulation structure using the finite element method, or by experimental measurement.

[0014] Preferably, in step S3, the heat balance equations for each node of the transient RC thermal network model are: ; in, Indicates the first 1 node For nodes heat capacity, For nodes temperature, For time, Represents nodes Adjacent nodes, For nodes The set of all adjacent nodes, For nodes With neighboring nodes The thermal resistance between them Adjacent nodes temperature, For nodes The heat source item at the location.

[0015] Preferably, in step S4, the multi-objective intelligent optimization algorithm is a decomposition-based multi-objective evolutionary algorithm. This algorithm uses the Chebyshev aggregation method to decompose the multi-objective optimization problem into several scalar quantum problems. Each sub-problem achieves the approximation of the Pareto front through cooperative evolution between adjacent sub-problems.

[0016] Preferably, in step S4, the thermal insulation performance index is the average temperature or heat flux density of the cold side of the thermal insulation structure predicted by the thermal network model, and the structural quality index is the total mass of the thermal insulation structure.

[0017] Preferably, in step S5, the preset decision criterion is a fuzzy compromise decision criterion. The optimal compromise solution is selected based on the relative proximity between each Pareto solution and the positive ideal solution and the negative ideal solution. The formula for calculating the relative proximity is: ; in, Indicates the first A Pareto solution, For the first The relative closeness of each Pareto solution For the first The weighted Euclidean distance between each Pareto solution and the positive ideal solution For the first The weighted Euclidean distance between the Pareto solution and the negative ideal solution.

[0018] Preferably, in step S4, the material density distribution within the design domain is described using a density field method based on spatial mapping, and the design variable is the pseudo-density of each finite element. The pseudo-density is sharpened using the Sigmoid projection function. ; in, Indicates the first Finite element elements, For the first The physical density of each unit For the first Intermediate design variables for each unit, This is the projection sharpness parameter. This is the projection threshold.

[0019] Therefore, the multi-objective optimization method for thermal insulation topology based on iterative correction of the above-mentioned structure using a fluid-structure interaction thermal network, as described in this invention, has the following beneficial effects: (1) This invention uses an adaptive iterative correction mechanism for the convective heat transfer coefficient and a high-fidelity reference temperature field to perform reverse calibration on the thermal network prediction model, which significantly improves the calculation accuracy of the control system state prediction model while maintaining its high calculation efficiency. This effectively overcomes the technical defects of the traditional thermal network control model, which relies on empirical formulas and causes deviations in the estimation of convective thermal resistance, and improves the robustness of the control system. (2) This invention introduces lumped heat capacity elements into the modified thermal network model to construct an RC network, which extends the steady-state prediction model into a transient state-space model, thereby realizing an accurate description of the transient heat transfer process of the system and providing an efficient calculation tool for dynamic thermal response prediction and feedforward control of the thermal insulation system. (3) The present invention embeds the modified high-precision thermal network model into the multi-objective control decision framework, and establishes a direct closed-loop link from state prediction to control strategy optimization. It can quickly evaluate the temperature field and transient response of any material distribution scheme or medium regulation parameter during the control decision process, and realize the fast and directional optimization control driven by system performance.

[0020] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0021] Figure 1 This is an overall flowchart of a multi-objective optimization method for thermal insulation topology based on iterative correction of fluid-structure interaction thermal networks according to the present invention. Figure 2 This is a comparison curve of the predicted temperature values ​​of the thermal network model before and after iterative correction with the reference temperature field in an embodiment of the present invention. Figure 3 This is a Pareto front distribution map obtained by multi-objective optimization in an embodiment of the present invention. Detailed Implementation

[0022] The technical solution of the present invention will be further described below with reference to the accompanying drawings and embodiments.

[0023] Unless otherwise defined, the technical or scientific terms used in this invention shall have the ordinary meaning understood by one of ordinary skill in the art to which this invention pertains. The terms "first," "second," and similar terms used in this invention do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word encompasses the elements or objects listed following the word and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0024] Example like Figure 1 As shown, this invention provides a multi-objective optimization method for thermal insulation topology based on iterative correction of fluid-structure interaction thermal networks. The technical solution of this invention will be described in detail below with reference to a specific thermal insulation structure design example: Step S1: Initial thermal network model construction: In this embodiment, the target thermal insulation structure is a one-dimensional multi-layer flat plate thermal insulation structure, consisting of a high-temperature heat source layer, an intermediate thermal insulation layer, and a low-temperature cooling channel. The design thickness of the thermal insulation layer is... The temperature boundary condition on the high-temperature side is The temperature of the cooling fluid on the low-temperature side is The cooling fluid flow rate is .

[0025] According to step S1, the insulation layer is uniformly divided along its thickness direction into... Each control unit is a node temperature control unit. ( The average temperature is represented by . The thermal resistance between adjacent nodes is calculated according to Fourier's law of thermal conductivity: ; in, For the first The thermal resistance of each control unit For the first The thickness of each control unit For the first Thermal conductivity of the material of each control unit This represents the heat transfer area.

[0026] The convective thermal resistance between the high-temperature heat source layer and the first insulation unit node, and the convective thermal resistance between the 20th insulation unit node and the cooling fluid are respectively: ; ; in, and These are the convective thermal resistances on the high-temperature side and the low-temperature side, respectively. and These are the convective heat transfer coefficients for the high-temperature side and the low-temperature side, respectively. and These are the convective heat transfer areas on the high-temperature side and the low-temperature side, respectively.

[0027] convective heat transfer coefficient and The initial values ​​are estimated based on the Dittus-Boelter empirical formula: ; ; in, For Nusselt numbers, Let Reynolds number be 1. For Prandtl numbers, The thermal conductivity of the fluid, This refers to the feature size.

[0028] Step S2, Iterative correction of convective heat transfer coefficient: The temperature field calculated using the finite volume method with a fine mesh (the number of meshes is 50 times the number of control volumes in the thermal network model) is used as the high-fidelity reference temperature field. .

[0029] like Figure 1 As shown, the iterative correction process for the convective heat transfer coefficient is as follows: Initialize the iteration counter Set relaxation factor Regularization parameters Local temperature relative change threshold Mean square temperature relative change threshold .

[0030] In the In the next iteration, the convective heat transfer coefficient of each convective heat transfer node is corrected using equation (1): (1); in, For the first In the nth iteration Convection heat transfer coefficient of each convection heat transfer node; For the first In the nth iteration Convection heat transfer coefficient of each convection heat transfer node; For the first Reference temperature values ​​at each node; For the first In the nth iteration The model predicts the temperature value at each node; This is the relaxation factor, used to control the step size of iterative corrections; This is a regularization parameter used to prevent the denominator from being zero.

[0031] After each iteration, the Gauss-Seidel iterative method is used to solve for the nodal temperatures of the thermal network model, obtaining the updated model-predicted temperature field. .

[0032] Convergence is judged based on equations (2) and (3): (2); (3); in, In this embodiment, the total number of temperature monitoring nodes is [number missing]. ; The threshold for relative local temperature change is taken in this embodiment. ; The threshold value for the relative change in mean square temperature is taken in this embodiment. .

[0033] When both of the above convergence conditions are met simultaneously, the iteration terminates, and the corrected convective heat transfer coefficient distribution and the corrected thermal network model are obtained.

[0034] like Figure 2 As shown, in this embodiment, the iteration converges in step 8. The horizontal axis in the figure represents the node position (along the insulation layer thickness direction), and the vertical axis represents the temperature value (unit: K). Curve A represents the high-fidelity reference temperature field, curve B represents the temperature field predicted by the thermal network model before correction, and curve C represents the temperature field predicted by the thermal network model after correction. The maximum deviation between the temperature field predicted by the thermal network model before correction and the reference temperature field is approximately 34.7 K, and the maximum deviation after correction is reduced to 3.2 K, improving the accuracy by about an order of magnitude, thus verifying the effectiveness of the iterative correction method of this invention.

[0035] Step S3: Establishment of transient RC thermal network model: According to step S3, at each node of the corrected thermal network model A lumped heat capacity element is introduced at the location, with a heat capacity value of Based on the specific heat capacity of the material ,density and the corresponding control volume calculate: ; in, For nodes heat capacity, For nodes The specific heat capacity of the material, For nodes The density of the material, For nodes The corresponding control volume.

[0036] In this embodiment, the heat insulation layer material is an alumina fiber-reinforced aerogel composite material with a specific heat capacity of [missing information]. ,density Volume of each control unit ,in For heat transfer area, For the first The thickness of each control unit.

[0037] node The heat balance equation is: (4); in, Indicates the first 1 node For nodes heat capacity, For nodes temperature, For time, Represents nodes Adjacent nodes, For nodes The set of all adjacent nodes, For nodes With neighboring nodes The thermal resistance between them Adjacent nodes temperature, For nodes The heat source item at the location.

[0038] By discretizing equation (4) using the implicit Euler method, we obtain: ; in, For time step index, For the first Each time step node temperature, For the first Each time step node temperature, For time step, For the first Adjacent nodes at each time step temperature, For the first Each time step node The heat source item at the location.

[0039] After processing, we obtained information about the temperature of each node. The system of linear algebraic equations is solved efficiently using the tridiagonal matrix algorithm (Thomas' algorithm). In this embodiment, the time step is taken as... The total calculation time is .

[0040] Step S4, Multi-objective topology optimization: According to step S4, the modified thermal network model is embedded as a solver into a multi-objective optimization framework. The optimization problem is formulated as follows: ; in, The vector representing the physical density of all units; The objective function vector; The average temperature on the cold side of the insulation layer is used as an indicator of insulation performance. ; The total mass of the insulation layer is used as a structural quality indicator. ; in, This refers to the cold-side heat exchange area; Temperature distribution on the cold side; Density of solid materials; For the first The volume of each unit; Volume constraints based on the maximum allowable material usage; This represents the total number of units.

[0041] The design variables are described using a density field method based on spatial mapping, and the pseudo-density is sharpened using a Sigmoid projection function. (5); in, For the first The physical density of each unit; For the first The intermediate design variables for each unit have a range of values. ; This refers to the projection sharpness parameter; This is the projection threshold. In this embodiment, the initial threshold is... As optimization iterations proceed, the size gradually increases to , .

[0042] The multi-objective optimization algorithm employs a decomposition-based multi-objective evolutionary algorithm (MOEA / D). This algorithm decomposes the multi-objective optimization problem into Chebyshev aggregation methods. The problem of individual quantum quanta: ; in, For Chebyshev aggregation functions; For the first The weight vector of each subproblem satisfies and ; As a reference point, ; Let l be the l-th objective function.

[0043] The MOEA / D algorithm approximates the Pareto front through cooperative evolution among adjacent subproblems. In this embodiment, the number of subproblems... The population size is 200, the maximum number of iterations is 500, the crossover probability is 0.9, and the mutation probability is... .

[0044] In each generation of optimization iteration, for each individual in the population, the modified thermal network model must be called to predict the temperature field and calculate the transient response in order to evaluate the individual's fitness function value.

[0045] like Figure 3 As shown, after 500 iterations of optimization, the Pareto front clearly demonstrates the trade-off between thermal insulation performance (average temperature on the cold side) and structural mass. The horizontal axis represents the average temperature on the cold side (in K), and the vertical axis represents the structural mass (in kg). A lower average temperature on the cold side results in better thermal insulation performance, but requires more insulation material, leading to an increase in structural mass; conversely, reducing the amount of material used will lead to an increase in the average temperature on the cold side.

[0046] Step S5: Optimal Solution Selection According to step S5, the optimal compromise solution is selected from the Pareto optimal solution set using the fuzzy compromise decision criterion. First, fuzzy membership normalization is performed on each objective function value: ; in, For the first The Pareto solution is at the th . Membership degree on each objective; For the first The Pareto solution of the first... One objective function value; For all Pareto solutions, the th The maximum value of each objective; For all Pareto solutions, the th The minimum value of each objective; This represents the number of solutions in the Pareto optimal solution set.

[0047] Weighted Euclidean distance between each Pareto solution and the positive ideal solution and the weighted Euclidean distance between the ideal solution and the negative ideal solution They are respectively: ; ; in, For the first The weight of each objective; ; In this embodiment, we take... .

[0048] Relative closeness of each Pareto solution to the ideal solution for: (6); in, For the first The relative similarity of the Pareto solutions; For the first The weighted Euclidean distance between each Pareto solution and the positive ideal solution; For the first Weighted Euclidean distance between Pareto solutions and negative ideal solutions. Relative proximity. The larger the value, the closer the solution is to the ideal solution.

[0049] Select relative closeness The largest Pareto solution is taken as the optimal compromise solution. Compared with the initial uniform design, the optimized solution reduces the average temperature on the cold side from the initial 423.5 K to 396.8 K, a reduction of about 26.7 K, while reducing the insulation layer mass by about 18.7%, thus significantly improving the insulation performance.

[0050] The working principle of this invention is as follows: First, an initial fluid-structure interaction lumped-parameter thermal network model of the target thermal insulation structure is constructed based on the thermoelectric analogy principle. This model uses thermal resistance elements to represent the heat conduction and convection heat transfer paths, and node temperatures to represent the average temperature of each control volume. Then, using a high-fidelity reference temperature field as a benchmark, the thermal network model is calibrated through a reverse iterative correction mechanism of the convective heat transfer coefficient, making the model's predicted temperature field approximate the reference temperature field, thereby overcoming the accuracy deficiency of traditional thermal network methods due to their reliance on empirical formulas for convective thermal resistance. Based on the corrected steady-state thermal network model, an RC network model is established by introducing lumped heat capacity elements at each node, expanding the capability from steady-state analysis to transient analysis. Finally, the corrected high-precision transient thermal network model is embedded as a solver within a multi-objective topology optimization framework. With thermal insulation performance and structural quality as optimization objectives, the MOEA / D multi-objective evolutionary algorithm iteratively searches for the optimal material distribution configuration within the design domain, achieving performance-driven design of the thermal insulation structure.

[0051] Therefore, this invention employs a multi-objective optimization method for thermal insulation topology based on iterative correction of a fluid-structure interaction thermal network with the aforementioned structure. By introducing an iterative correction mechanism for the convective heat transfer coefficient based on a reference temperature field, the computational accuracy of the thermal network method is significantly improved while maintaining its high-efficiency computational advantage. By integrating heat capacity elements into the corrected thermal network model, an accurate description of the transient heat transfer process is achieved. By embedding a high-precision thermal network model into a multi-objective optimization framework, a direct optimization link from thermal analysis to structural design is established, providing an efficient and reliable technical solution for performance-driven design of thermal insulation structures.

[0052] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A multi-objective optimization method for thermal insulation topology based on iterative correction of fluid-structure interaction thermal networks, characterized in that, Includes the following steps: S1: Construct an initial fluid-structure interaction thermal network model. The initial value of the convective heat transfer coefficient at the solid-fluid interface in the thermal network model is estimated based on an empirical formula. S2: Using the reference temperature field as a benchmark, the convective heat transfer coefficient is iteratively corrected in reverse until the temperature field predicted by the thermal network model and the reference temperature field are consistent under the preset convergence criterion. S3: Introduce lumped heat capacity elements at the nodes of the modified thermal network model to establish a transient RC thermal network model; S4: The modified thermal network model is embedded into a multi-objective optimization framework. The thermal insulation performance index and structural quality index are used as optimization objectives, and the material density distribution in the design domain is used as the design variable. A multi-objective intelligent optimization algorithm is used to iteratively find the optimal Pareto solution set. S5: Select the optimal compromise solution from the Pareto optimal solution set according to the preset decision criteria, and output it as the final thermal insulation structure topology optimization scheme.

2. The multi-objective optimization method for thermal insulation topology based on iterative correction of fluid-structure interaction thermal networks according to claim 1, characterized in that: In step S1, the initial fluid-structure interaction thermal network model is constructed as follows: the insulation structure is divided into multiple control volume units along the thickness direction. The average temperature within each control volume unit is characterized by the node temperature. Adjacent nodes are connected by conductive thermal resistance, and the solid-fluid interface is connected by convective thermal resistance. The expression for the convective thermal resistance is: ； in, For convective thermal resistance, The convective heat transfer coefficient is... This represents the convective heat transfer area.

3. The multi-objective optimization method for thermal insulation topology based on iterative correction of fluid-structure interaction thermal networks according to claim 1, characterized in that: In step S2, the reverse iterative correction of the convective heat transfer coefficient is as follows: in the first step... In the next iteration, the convective heat transfer coefficient of each convective heat transfer node is corrected according to the following formula: ； in, Indicates the first One convection heat transfer node For the first In the nth iteration The convective heat transfer coefficient of each node, For the first In the nth iteration The convective heat transfer coefficient of each node, For the first Reference temperature values ​​at each node For the first In the nth iteration The model predicts the temperature value at each node. As a relaxation factor, This is the regularization parameter.

4. The multi-objective optimization method for thermal insulation topology based on iterative correction of fluid-structure interaction thermal networks according to claim 1, characterized in that: In step S2, the preset convergence criterion is: the maximum value of the relative change in temperature of each node in two adjacent iterations is less than or equal to the first threshold. Meanwhile, the mean square value of the relative change in temperature at each node between two adjacent iterations is less than or equal to the second threshold. ,Right now: ； and ； in, This represents the total number of temperature monitoring nodes. For the first In the nth iteration The model predicts the temperature value at each node. This is the threshold for the relative change in local temperature. This represents the threshold for relative change in mean square temperature.

5. The multi-objective optimization method for thermal insulation topology based on iterative correction of fluid-structure interaction thermal networks according to claim 1, characterized in that: In step S2, the reference temperature field is obtained by any of the following methods: by calculating the target insulation structure after fine meshing using the finite volume method, or by numerical simulation of the target insulation structure using the finite element method, or by experimental measurement.

6. The multi-objective optimization method for thermal insulation topology based on iterative correction of fluid-structure interaction thermal networks according to claim 1, characterized in that: In step S3, the heat balance equations for each node of the transient RC thermal network model are: ； in, Indicates the first 1 node For nodes heat capacity value, For nodes temperature, For time, Represents nodes Adjacent nodes, For nodes The set of all adjacent nodes, For nodes With neighboring nodes The thermal resistance between them Adjacent nodes temperature, For nodes The heat source item at the location.

7. The multi-objective optimization method for thermal insulation topology based on iterative correction of fluid-structure interaction thermal networks according to claim 1, characterized in that: In step S4, the multi-objective intelligent optimization algorithm is a decomposition-based multi-objective evolutionary algorithm. This algorithm uses the Chebyshev aggregation method to decompose the multi-objective optimization problem into several scalar quantum problems. Each sub-problem achieves the approximation of the Pareto front through cooperative evolution between adjacent sub-problems.

8. The multi-objective optimization method for thermal insulation topology based on iterative correction of fluid-structure interaction thermal networks according to claim 1, characterized in that: In step S4, the thermal insulation performance index is the average temperature or heat flux density of the cold side of the thermal insulation structure predicted by the thermal network model, and the structural quality index is the total mass of the thermal insulation structure.

9. The multi-objective optimization method for thermal insulation topology based on iterative correction of fluid-structure interaction thermal networks according to claim 1, characterized in that: In step S5, the preset decision criterion is a fuzzy compromise decision criterion. The optimal compromise solution is selected based on the relative proximity between each Pareto solution and the positive ideal solution and the negative ideal solution. The formula for calculating the relative proximity is: ； in, Indicates the first A Pareto solution, For the first The relative closeness of each Pareto solution For the first The weighted Euclidean distance between each Pareto solution and the positive ideal solution For the first The weighted Euclidean distance between the Pareto solution and the negative ideal solution.

10. The multi-objective optimization method for thermal insulation topology based on iterative correction of fluid-structure interaction thermal networks according to claim 1, characterized in that: In step S4, the material density distribution within the design domain is described using a density field method based on spatial mapping. The design variable is the pseudo-density of each finite element, and the sharpness of the pseudo-density is controlled by the Sigmoid projection function. ； in, Indicates the first Finite element elements, For the first The physical density of each unit For the first Intermediate design variables for each unit, This is the projection sharpness parameter. This is the projection threshold.

Images