A method and system for designing an ultra-high voltage insulator based on multi-field coupling

Abstract

A kind of design optimization method and system of ultra-high voltage insulator based on multi-field coupling.Establish the dynamic correlation model of insulator material parameters changing with temperature and field intensity, and construct the multi-physical field coupling model containing the interaction of electric field, thermal field, mechanical field and flow field. Through multi-field joint solution and cyclic iteration, the field distribution data under stable state is obtained;Based on the electrical, thermal and mechanical failure risk indicators, a coupled risk objective function is constructed, a multi-objective optimization model is established with structural parameters as optimization variables, and a double-path strategy combining single-objective optimization and Pareto frontier solution is used for iterative optimization. At the same time, a surrogate model is constructed between the optimization variables and the risk indicators, and the error correction and model updating are carried out according to the solution results, and the insulator optimization design scheme meeting the multi-failure constraints is output. The present application can improve the reliability of the optimization design of ultra-high voltage insulator.

Description

Technical Field

[0001] This invention belongs to the field of ultra-high voltage insulator technology, specifically relating to an ultra-high voltage insulator design optimization method and system based on multi-field coupling. Background Technology

[0002] Under the long-term synergistic effect of strong electric fields, complex mechanical loads and harsh natural environments, the failure modes of insulators exhibit multi-physics coupling characteristics, mainly manifested as electrical breakdown, thermal brittle fracture and mechanical fatigue.

[0003] Currently, insulator design typically focuses on optimizing the electric field distribution to control corona and flashover; to ensure mechanical strength meets safety factors, thermal analysis is often used as a post-hoc verification or relies solely on empirical formulas. This approach ignores the strong mutual influence and coupling between the electric field, thermal field, and mechanical stress field, which may result in unknown weaknesses in the designed insulators under real multi-field synergy.

[0004] Furthermore, in existing technologies, engineers may use general-purpose finite element software to simulate a single field (such as electrostatic field analysis only), or perform simple one-way data transfer between two fields (such as using electric field calculation results as a heat source for steady-state temperature analysis). These methods cannot achieve real-time, bidirectional, and fully coupled simulation of multiple physical fields such as electro-thermal-mechanical-fluid fields. Their optimization objectives are often singular, such as "lowest field strength" or "lightest weight," lacking a multi-objective collaborative optimization framework that comprehensively balances electrical performance, thermal stability, mechanical reliability, and even economic efficiency. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a method and system for optimizing the design of ultra-high voltage insulators based on multi-field coupling. This invention acquires multi-field response data of ultra-high voltage insulators under operating conditions by constructing coupled models of electric, thermal, mechanical, and flow fields. A multi-objective optimization model is established based on multiple failure risk indicators, and combined with dual-path optimization and iterative updates using a surrogate model, the collaborative optimization of insulator structural parameters is achieved. Multi-field coupling analysis reduces failure risks during the design process and improves the reliability of insulator design.

[0006] The first aspect of this application discloses a design optimization method for ultra-high voltage insulators based on multi-field coupling, which adopts the following technical solution: Obtain the structural and material parameters of the composite insulator; define the set of structural parameter variables and the set of material parameters, and establish a dynamic correlation model between material parameters and temperature and field strength; establish multi-field data interaction relationships to model the multi-physics coupling parameter space; Based on the coupled parameter space, a multi-field joint solution model is established; the electric field calculation transfers the heat loss source to the thermal field, the thermal field calculation feeds back material parameters to the electric field and transfers temperature loads to the mechanical field; the mechanical field calculation feeds back structural deformation information to the electric field and the flow field, and the flow field calculation feeds back heat transfer boundary conditions to the thermal field; multi-field cyclic iteration is performed to output multi-field data under steady state; Based on the multi-field data, risk indicators for electrical failure, thermal failure, and mechanical failure are calculated respectively. Coupled risk objectives are introduced, and a multi-objective optimization function is constructed with structural parameter variables as optimization variables. A dual-path solution strategy is constructed, including transforming the multi-objective optimization function into a single objective to obtain a single optimal solution, and obtaining the Pareto front of the multi-objective. A multi-objective surrogate model for solving optimization variables and risk indicators is established, and a dual-path iterative solution is performed. During the iteration process, the solution results of the dual paths are interactively verified, and the candidate solutions are fed back to the surrogate model for error calculation to update the training of the surrogate model until both paths converge and the optimal solution is output.

[0007] Furthermore, the definition process of the set of structural parameter variables includes: In the simulation analysis environment, the structural parameters of the composite insulator are constructed into a three-dimensional geometric model through the corresponding simulation mapping method; the three-dimensional geometric model is parameterized, and the optimizable structural parameters are defined as a set of structural parameter variables. Set corresponding physical constraint boundaries and coupling constraints for each variable in the set of structural parameter variables.

[0008] Furthermore, the multiphysics coupling parameter space includes a set of structural parameter variables, a set of material parameters, a set of environmental parameters, and a dynamic correlation model of material parameters as a function of temperature and electric field. The dynamic correlation models of material parameters with temperature and electric field include temperature correlation models of relative permittivity, electrical conductivity, elastic modulus, and thermal expansion.

[0009] Furthermore, the multi-field joint solution model is used to perform data interaction between different physical fields, including: establishing an electric field model to calculate the electric field strength, combining the leakage current and the relative permittivity of the material to calculate the leakage current heat source and the dielectric loss heat source, forming a total loss heat source that is transferred to the thermal field calculation; A thermal field model is established to calculate heat conduction based on the total heat loss source. Convection and radiation heat transfer boundaries are set, and the temperature distribution of the composite insulator is obtained by solving. The relative permittivity, conductivity and thermal expansion coefficient of the calculated material are updated based on the temperature distribution and fed back to the electric field model.

[0010] The thermal field model transfers the temperature distribution of the insulator to the mechanical field model; The mechanical field model calculates the thermal strain and structural displacement of the material based on the temperature distribution and the mechanical load during the operation of the insulator, updates the three-dimensional geometric model of the insulator, and transfers the thermal strain and structural displacement to the flow field model; A flow field model is established to calculate the velocity distribution and surface convection heat transfer coefficient based on the operating environment parameters, and then fed back to the thermal field model to update the heat transfer boundary conditions.

[0011] Furthermore, the multi-field cyclic iteration includes: Based on the data interaction of the electric field model, thermal field model, mechanical field model, and flow field model, the comprehensive coupling error of each iteration is calculated; when the comprehensive coupling error is less than the convergence threshold, the multiphysics data of the current iteration is determined to be multiphysics data output in a steady state. The multi-field data under steady-state conditions include electric field distribution, temperature distribution, and stress distribution.

[0012] Furthermore, the calculation process of the comprehensive risk index includes: For electrical failure risk indicators, the maximum field strength, maximum field strength gradient, and area of ​​the concentrated electric field region are extracted from the electric field distribution in the multi-field data. For the thermal failure risk index, it is calculated based on the highest temperature, temperature rise rate and high temperature duration extracted from the temperature distribution in the multi-field data. For mechanical failure risk indicators, mechanical stress, maximum displacement, and fatigue damage are calculated based on the stress distribution in the data.

[0013] Furthermore, the multi-objective optimization function includes: Minimize electrical failure risk indicators, thermal failure risk indicators, mechanical failure risk indicators, and manufacturing costs; it also includes coupled risk objectives, which aim to minimize coupled impact factors; The coupling influence factor is calculated based on the fusion of electrical failure risk indicators, thermal failure risk indicators, and mechanical failure risk indicators.

[0014] Furthermore, the transformation of the multi-objective optimization function into a single-objective solution for obtaining a single optimal solution includes: The electrical failure risk index, thermal failure risk index, mechanical failure risk index, manufacturing cost, and coupling influencing factors are weighted to obtain a single-objective optimization function; the weight of each objective item is dynamically adjusted according to the operating conditions.

[0015] Furthermore, the interactive verification of the solution results for the two paths during the iteration process includes: Project the current optimal solution output by the single-objective path onto the Pareto front of the multi-objective path to determine the dominance relationship; If the current optimal solution is not dominated by any solution in the Pareto front, continue the current iteration; If the current optimal solution is dominated by at least one solution in the Pareto front, the improvement of the dominant solution relative to the current optimal solution in each objective term is calculated, and the weight of each objective term in the single objective path is adjusted according to the improvement.

[0016] The second aspect of this application discloses a system for optimizing the design of ultra-high voltage insulators based on multi-field coupling, which implements the optimization method described in the first aspect of this application. The system includes: The coupling space establishment module acquires the structural and material parameters of the composite insulator; defines the set of structural parameter variables and the set of material parameters, and establishes a dynamic correlation model between material parameters and temperature and field strength; and establishes multi-field data interaction relationships to model the multi-physics coupling parameter space. Multi-field coupled solution module; based on the coupled parameter space, a multi-field joint solution model is established; electric field calculation transmits the heat loss source to the thermal field, thermal field calculation feeds back material parameters to the electric field and transmits temperature load to the mechanical field; mechanical field calculation feeds back structural deformation information to the electric field and flow field, flow field calculation feeds back heat transfer boundary conditions to the thermal field; executes multi-field cyclic iteration and outputs multi-field data under steady state; A multi-objective optimization module is used to calculate risk indicators for electrical failure, thermal failure, and mechanical failure based on the multi-field data, introduce coupled risk objectives, and construct a multi-objective optimization function with structural parameter variables as optimization variables. A dual-path solution strategy is constructed, including transforming the multi-objective optimization function into a single objective to obtain a single optimal solution, and obtaining the Pareto front of the multi-objective. The optimization solution module establishes a multi-objective surrogate model for solving optimization variables and risk indicators, and performs dual-path iterative solutions. During the iteration process, the solution results of the dual paths are interactively verified, and candidate solutions are fed back to the surrogate model for error calculation to update the training of the surrogate model until both paths converge and the optimal solution is output.

[0017] The beneficial effects of this invention are that, compared with the prior art, 1. By constructing a two-way data interaction mechanism among electric field, thermal field, mechanical field, and flow field, and introducing a dynamic correlation model of material parameters changing with temperature and electric field, multi-physics coupling modeling and joint solution of composite insulators under complex operating environments are realized. This invention comprehensively reflects the interrelationships between electric field distribution, temperature change, structural deformation, and stress response, avoiding evaluation biases caused by single-physics field analysis and improving the accuracy of insulator operating status analysis and failure risk prediction.

[0018] 2. A multi-objective optimization model covering electrical failure, thermal failure, mechanical failure, and manufacturing cost is constructed based on multi-field calculation results. A dual-path solution strategy combining single-objective optimization and Pareto front search is adopted to achieve comprehensive optimization of insulator structural parameters. Simultaneously, by establishing a surrogate model between optimization variables and risk indicators, and utilizing a dual-path result interaction verification and iterative update mechanism, redundant solutions in the complex multi-field calculation process are reduced, improving the convergence efficiency of the optimization process. This ensures that the obtained design scheme can reduce multiple types of failure risks while also meeting engineering manufacturing requirements. Attached Figure Description

[0019] Figure 1 This is a flowchart illustrating the insulator design optimization method provided in this embodiment. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of this invention. The embodiments described in this application are merely some embodiments of this invention, and not all embodiments. Based on the spirit of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the protection scope of this invention.

[0021] Example 1 As an embodiment of this application, a specific implementation method for an ultra-high voltage insulator design optimization method based on multi-field coupling is disclosed. The execution flow of the method embodiment is as follows: Figure 1 .

[0022] S1. In one embodiment, the structural parameters and material parameters of the composite insulator are obtained; the set of structural parameter variables and the set of material parameters are defined, and a dynamic correlation model between material parameters and temperature and field strength is established; a multi-field data interaction relationship is established to model the multi-physics field coupling parameter space.

[0023] Obtain the structural parameters, material parameters, and operating environment parameters of composite insulators, and construct a design parameter space for multi-physics coupling analysis.

[0024] 1.1 Obtain the structural parameters of the composite insulator; Insulator structural parameters include: number of sheds , Apron diameter (outer diameter) Umbrella skirt spacing Sheath thickness mandrel dimensions (diameter) Equalizing ring dimensions (radius) Pipe diameter ) and installation location (installation height) ); The above structural parameters constitute the three-dimensional geometric model of the insulator. In further explanation, the function It is not a specific mathematical function, but rather a parametric construction mapping used to represent the three-dimensional geometric model of a composite insulator. Its input is structural parameters, and its output is a three-dimensional geometric entity that can be used for finite element analysis. In actual engineering implementation, It corresponds to the parametric modeling script in simulation software.

[0025] Furthermore, the ensemble model is parameterized by defining the optimizable structural parameters as a set of structural parameter variables: This set of variables is used by subsequent optimization algorithms to automatically adjust the structural parameters.

[0026] In a specific, referable manner, the set of structural parameter variables This constitutes a feasible region in N-dimensional real space. To ensure the generated geometric solid model... It is feasible in terms of engineering and meets the requirements for electrical and mechanical insulation. The variables in the code define physical constraint boundaries and coupling constraints: Set the number of umbrella skirts ; For the set of integers, and These are the minimum and maximum possible values ​​for the set number of umbrella skirts; Set the outer diameter of the umbrella skirt ; This represents the minimum possible value for the set mandrel diameter. For the thickness of the sheath, This represents the maximum possible value for the outer diameter of the umbrella skirt.

[0027] Set the spacing of the umbrella skirts And it satisfies the creepage distance constraint; and These are the minimum and maximum possible values ​​for the set skirt spacing, respectively; Sheath thickness Equalizing ring dimensions (radius) Pipe diameter ) and installation location (installation height) All of these take values ​​continuously within the range of 0 and their respective maximum values.

[0028] If the optimization algorithm generates parameters that violate geometric interference, such as the equalizing ring contacting the umbrella skirt, then a penalty function is introduced. This makes it infeasible in the corresponding iteration step.

[0029] In a specific, referable approach, the mapping function is constructed using parameterization. Specifically, this is implemented as a parametric set reconstruction interface for simulation software. It defines the input structure parameters. With geometric solid model The mapping relationship is as follows: ; This refers to pre-written macro commands or Python interface scripts. When the optimization algorithm outputs a new set of... At that time, the script is automatically invoked to update node coordinates and trigger adaptive mesh generation to ensure... One-to-one correspondence with finite element meshes.

[0030] 1.2 Obtain the material parameters of the insulator; Insulator material parameters include: relative permittivity Electrical conductivity Thermal conductivity ,density Specific heat capacity Elastic modulus Poisson's ratio Coefficient of thermal expansion ; Forming a set of material parameters Different material regions are assigned different physical property parameters, including the silicone rubber sheath region, the glass fiber core rod region, and the metal equalizing ring region.

[0031] 1.3 Obtain runtime environment parameters; Operating environment parameters include: operating voltage Ambient temperature ,humidity Wind speed filth level (Quantified in engineering implementation using equivalent salt density ESDD and ash density NSDD), convective heat transfer coefficient ; Forming a set of environmental parameters Among them, the convective heat transfer coefficient of the insulator surface According to wind speed Calculations are used to set thermal field boundary conditions.

[0032] In one possible implementation, runtime environment parameters Instead of a random walk, several typical operating scenarios are defined based on the actual operational risks of the power system. For example: Operating Condition 1: Conventional drying (voltage is rated voltage) ): ; Operating Condition 2: Heavy pollution, high temperature, and quiet air (voltage is rated voltage) ): ; Operating Condition 3: Low temperature, high humidity, and strong wind (voltage is rated voltage) ): .

[0033] 1.4 Establish a dynamic correlation model of material parameters as a function of temperature and electric field; Establish a dynamic correlation model of material parameters as a function of temperature and electric field intensity. This includes the relationship between dielectric parameters and temperature, the relationship between conductivity and temperature rise, and the relationship between elastic modulus and temperature. Specifically: (1) Taking into account the effect of temperature on dielectric properties, a temperature correlation model for relative permittivity is established: ; in, Indicates temperature as The relative permittivity at time , Represents the initial dielectric constant. This represents the dielectric temperature coefficient. A dielectric constant-temperature correlation model is used to feed back the thermal field results to the electric field model. Furthermore, the initial dielectric constant represents the relative dielectric constant of the composite insulator material under reference conditions (temperature 20 / 25℃, initial electric field strength, dry and uncontaminated state). Typically, the initial dielectric constant of a silicone rubber sheath is 2-3, and that of an epoxy glass fiber core rod is 3-4.

[0034] In a further embodiment, in the multiphysics simulation of composite insulators, the node temperature variation coefficient is typically segmented based on the glass transition temperature of the material and the actual operating temperature rise range. For example: In the low-temperature region (20°C below the glass transition temperature), silicone rubber is in a glassy state, and the dielectric temperature coefficient is basically stable or rises slowly, so a small positive value can be taken. In the high-temperature region (20°C above the glass transition temperature), silicone rubber is in a highly elastic state, and the dielectric temperature coefficient tends to level off or decrease slightly, so a small positive value (less than the value in the low-temperature region) or a negative value can be taken. In the transition region (between the upper limit of the low-temperature region and the lower limit of the high-temperature region), silicone rubber is in the glass transition region, and the dielectric temperature coefficient will rise sharply or fluctuate, exhibiting nonlinearity, so the average slope or the fitted slope can be taken.

[0035] (2) Based on the thermal activation characteristics of the material, establish a conductivity-temperature correlation model: ; in, Indicates temperature The electrical conductivity of the material at that time Indicates the initial conductivity. Indicates the material activation energy. This represents the Boltzmann constant. The process of leakage current change due to temperature rise is described using a conductivity-temperature correlation model.

[0036] In a further implementation, the initial conductivity is the volume conductivity of the insulator material under reference conditions, which can be designed based on engineering experience or test data; in actual implementation through engineering simulation, the initial value can be taken as... The activation energy of a material reflects the sensitivity of its conductivity to temperature and can be obtained by fitting the slope of an Arrhenius curve. For example, the activation energy of conductivity for polymer insulating materials such as silicone rubber is typically... between.

[0037] (3) Considering the decrease in material mechanical properties due to high temperature, a temperature correlation model for elastic modulus is established: ; in, Indicates temperature The elastic modulus at that time, This represents the initial elastic modulus. This represents the elastic modulus attenuation coefficient. This temperature-dependent model of elastic modulus is used for mechanical stress calculations.

[0038] In a further embodiment, the initial elastic modulus is the Young's modulus of the insulator material under reference conditions; the typical range for parts such as silicone rubber sheath and skirts is between 0.5 and 5.0 MPa, and the typical range for parts such as glass fiber resin core rod is 10-30 GPa.

[0039] (4) Based on the relationship between temperature change and structural deformation, a thermal expansion model is established: ; in, This indicates the amount of thermal deformation of the structure. Indicates the coefficient of thermal expansion. This represents the initial dimensions. The thermal expansion model is used for subsequent calculations of the mechanical field.

[0040] In one possible implementation, to achieve automatic attribute updates in multiphysics iterations, a dynamic correlation model can be used. Discretize into a lookupable numerical model or a piecewise function library. In an example using the material silicone rubber: Temperature correlation of relative permittivity:

[0041] These respectively indicate that silicone rubber in the low-temperature region ( Less than 20°C below the glass transition temperature Glassy state (dielectric temperature change coefficient) =0.001 / ℃) below, transition region (between the upper limit of the low temperature region and the lower limit of the high temperature region) glass transition region (dielectric temperature change coefficient) =0.01 / ℃, This is the transition compensation constant, used to maintain the piecewise function at... (continuity at the location), and high temperature zone ( Greater than 20°C above the glass transition temperature Highly elastic state (dielectric temperature change coefficient) The relative permittivity is -0.0005 / ℃.

[0042] For silicone rubber sheaths, setting the activation energy initial conductivity Boltzmann constant In the solver, the temperature is calculated in real time. Substitute the conductivity-temperature correlation model into the unit material to update its conductivity properties.

[0043] 1.5. Based on the three-dimensional geometric model, the insulator is meshed using finite element methods, and the computational domain is divided according to different physical processes. This includes the electric field domain, the heat conduction domain, the structural stress domain, and the fluid heat transfer domain.

[0044] Establish data interaction relationships between various physical fields: the dielectric loss and leakage current loss in the electric field calculation results are used as the heat source input of the thermal field; the temperature distribution obtained from the thermal field calculation is used to update the dielectric constant, conductivity and mechanical parameters of the material; the mechanical field calculates the structural deformation based on the thermal strain generated by the temperature rise, and feeds the deformation results back to the geometric model to update the electric field calculation domain; the flow field calculates the convective heat transfer coefficient based on the ambient wind speed and temperature, and feeds it back to the thermal field boundary conditions, realizing bidirectional coupled modeling of multiple physical fields of electric-thermal-mechanical-fluid.

[0045] (1) Electric field calculation region; Define the electric field region as ,satisfy ;in, For electric potential, It is the relative permittivity, used to calculate space potential and electric field strength. .

[0046] (2) Heat conduction region; Define the heat conduction region as The heat conduction equation is established as follows: ;in, It is a heat source of loss generated by the electric field.

[0047] (3) Structural stress region; Define the structural stress region as The structural mechanics equations are established as follows: ;in, Mechanical stress, For external mechanical loads.

[0048] (4) Fluid heat transfer zone; When a natural air cooling or active cooling structure is present, a fluid zone is established. The continuity equations used in fluid control and energy equation Calculate fluid flow and heat transfer processes.

[0049] 1.6 Finally, the structural parameters, material dynamic parameters, and environmental parameters are integrated to form a multiphysics coupled parameter space, represented as: ; Represents dynamic material parameters that vary with temperature and electric field.

[0050] By establishing a multi-dimensional design parameter space that includes structural parameters, material dynamic parameters, and environmental parameters, and introducing a correlation model of material properties changing with temperature and electric field, subsequent simulations of electric field, thermal field, mechanical field, and flow field can achieve bidirectional data interaction based on a unified parameter system, avoiding the problem of insufficient failure risk assessment caused by fixed material property parameters and single-field analysis in traditional design processes.

[0051] S2. In one implementation, a multi-field joint solution model is established based on the coupled parameter space; the electric field calculation transmits the heat loss source to the thermal field, the thermal field calculation feeds back material parameters to the electric field and transmits temperature loads to the mechanical field; the mechanical field calculation feeds back structural deformation information to the electric field and the flow field, and the flow field calculation feeds back heat transfer boundary conditions to the thermal field; multi-field cyclic iteration is performed to output multi-field data under steady-state conditions.

[0052] In S1, the data interaction relationships between the various physical fields can be described as follows: the electric field transfers electrical loss heat sources to the thermal field; the thermal field feeds back temperature-related material parameters to the electric field; the thermal field transfers temperature loads to the mechanical field; the mechanical field feeds back structural deformation information to the electric and flow fields; and the flow field feeds back heat transfer boundary conditions to the thermal field. Multiphysics iterative calculations are achieved through these data interactions.

[0053] 2.1 Electric field model calculation; Based on the electrode structure, applied voltage, and material dielectric parameters of the composite insulator, a finite element method (FEM) model for calculating the electric field is established. The governing equations for the electric field are adopted from Maxwell's equations, specifically the electrostatic field governing equations. ,in. The relative permittivity of the material is . It represents the electric potential in space.

[0054] Furthermore, the potential boundary conditions are set according to the high-voltage end and grounding end of the insulator as follows: and ;in, Operating voltage and These are the high-voltage electrode boundary and the grounding boundary, respectively.

[0055] The spatial potential distribution is obtained by solving the problem. Further calculation of the space electric field strength To obtain the maximum electric field strength And regions of concentrated electric field. Among them, regions of concentrated electric field. ; This is the preset electric field strength threshold.

[0056] Furthermore, an electric loss heat source model is established based on the electric field calculation results, and the electric field energy loss is converted into a thermal field input. The dielectric loss heat source is represented as follows: , As a heat source for dielectric loss, The electric field strength in space, Angular frequency, This is the dielectric loss angle. and These are the initial permittivity and the relative permittivity, respectively.

[0057] For areas with surface contamination or leakage current, calculate the heat source of the leakage current. ; Leakage current density, is the electrical conductivity.

[0058] The total heat loss source .Will Transferred to the thermal field calculation area.

[0059] 2.2 Thermal field model calculation; Based on the heat loss source obtained from the electric field calculation, a heat conduction model of the insulator is established; the governing equation of the thermal field is: ; in, For material density, For specific heat capacity, Thermal conductivity, Indicates the temperature of the insulator. The heat source of loss is calculated and output by the electric field model.

[0060] For the heat transfer process between the insulator surface and the ambient air, a convective heat transfer boundary is set: ; in, The convective heat transfer coefficient is... The ambient temperature.

[0061] For high-temperature radiation regions, a radiation heat transfer model is introduced: ; Indicates emissivity. is the Stefan-Boltzmann constant. In one possible implementation, the high-temperature radiation region is defined by combining the set and the operating condition, exemplarily: the surface of the equalizing ring, the junction of the umbrella skirt root and the sheath, and the entire surface under special conditions (heavily contaminated or at an ambient temperature greater than 35°C).

[0062] In one possible implementation, during the iteration, the radiative heat transfer model is applied as a Neumann boundary to the outer surface of the insulator, and its finite element discretization form is as follows: ; in, It is the surface outward normal vector; The equivalent radiative heat transfer coefficient is obtained through linearization: ; denoted as surface emissivity (dimensionless).

[0063] In each iteration, this boundary condition recalculates the equivalent radiative heat transfer coefficient based on the current surface temperature, thus simulating nonlinear radiative heat transfer.

[0064] The temperature distribution is obtained by solving the thermal field. Maximum temperature and temperature gradient .

[0065] Furthermore, the material properties, including relative permittivity, electrical conductivity, and coefficient of thermal expansion, are updated based on the thermal field calculation results. The electric field model is then re-inputted to correct the electric field calculations for the next round. This achieves feedback between the thermal field and the electric field.

[0066] 2.3 Calculation of the mechanical field model; Based on the temperature distribution obtained from thermal field calculations and the mechanical loads during insulator operation, a structural mechanics calculation model is established. The governing equations of the mechanical field are as follows: ;in, Mechanical stress, External mechanical load; Considering the thermal expansion effect caused by temperature, establish the thermal strain. , For mechanical strain, the total strain relationship is: According to the constitutive relation of materials , Given the material elasticity matrix, the equivalent stress can be calculated. and structural displacement .

[0067] Furthermore, the structural displacement updates the set of structural parameter variables of the insulator. The three-dimensional geometric model of the deformed insulator is then... The updated electrode distance, skirt position, and equalizing ring position are then re-input into the electric field calculation model. Simultaneously, the deformed outer boundary is input into the fluid model to update the fluid flow region.

[0068] In a specific implementation that can be referenced, in order to realize the transfer of structural deformation to the flow field region, an arbitrary Lagrange-Eulerian method can be used to handle the mesh deformation of the flow field region, for example: Structural displacement calculated from mechanical field The mesh is mapped to the boundary nodes of the fluid domain using radial basis function interpolation; for mesh nodes inside the fluid domain, the Laplace smoothing method can be used for network deformation.

[0069] 2.4 Flow field model calculation; A fluid heat transfer model is established based on the natural wind field or active cooling structure in the insulator's operating environment. This includes the continuity equation. Momentum equation and the energy equation . Indicates wind speed. For pressure, Indicates dynamic viscosity.

[0070] The velocity distribution, temperature boundary layer, and surface convective heat transfer coefficient are obtained by solving the flow field. The surface convective heat transfer coefficient is calculated as follows: ; For Nusselt numbers, The thermal conductivity of the fluid, is the thermal conductivity of the fluid.

[0071] Will Convection heat transfer boundary fed back to the thermal field model This enables feedback from the flow field to the thermal field.

[0072] 2.5. Establish a cyclical iterative mechanism for the electric field, thermal field, mechanical field, and flow field. The iteration process is represented as follows: Calculate the electric field based on the current material parameters and structural model. ,according to Calculate heat loss source Further solve the thermal field model to output the temperature. ,according to Update relative permittivity, conductivity, and mechanical deformation. This process updates the three-dimensional geometric model of the insulator. Finally, the convective heat transfer coefficient in the flow field is calculated based on the updated structural parameters and temperature.

[0073] Execute the loop until the convergence condition is met. ; To synthesize the coupled residuals, Let be the convergence threshold. Further, the integrated coupling residual is expressed as: ; in, , and The weighting coefficients for the residual terms can be set to 0.5, 0.3, and 0.2 for example during the initial iteration. Once the convergence condition is met, the electric field distribution, temperature distribution, stress distribution, and fluid heat transfer state under steady-state conditions are output as inputs to the subsequent multi-objective optimization model.

[0074] S3. In one implementation, based on the coupled calculation results of electric field, thermal field, mechanical field, and flow field obtained in S2, a multi-physics failure risk assessment model for the long-term operational reliability of composite insulators is constructed. This model extracts electrical failure characteristic parameters, thermal failure characteristic parameters, and mechanical failure characteristic parameters to establish corresponding failure risk assessment sub-models, and forms a comprehensive risk assessment index through multi-factor fusion.

[0075] 3.1. Based on the electric field calculation results, extract the maximum electric field strength, electric field gradient, and area of ​​the electric field concentration region to establish an electrical failure risk assessment model.

[0076] Electrical failures mainly include corona discharge caused by excessively high electric field strength and localized insulation breakdown caused by concentrated electric field; this can be defined as an electrical failure risk indicator: ; in, Indicates the maximum electric field strength. Indicates the maximum electric field gradient. Indicates the area of ​​the region where the electric field is concentrated; The function for calculating electrical risk indicators is as follows: Define the maximum field strength normalization index as , The maximum allowable electric field strength for the material or structure; when If the electric field strength exceeds the safety standard, it means that the field strength exceeds the safety standard.

[0077] Define the field strength gradient risk normalization index as: , To allow the maximum electric field gradient; maximum electric field gradient electric field gradient The electric field gradient reflects the degree of change in local field strength.

[0078] Define the risk index of the electric field concentration area as ; This represents the total surface area of ​​the insulator. The larger this ratio, the higher the proportion of high field strength regions, and the higher the probability of partial discharge and electrical tree aging.

[0079] Furthermore, the risk of integrated electrical failures , , and The weight coefficients for each indicator item satisfy the following conditions: .

[0080] 3.2. Based on the temperature field results obtained in S2, establish a thermal aging risk assessment model.

[0081] Thermal failure primarily stems from prolonged high temperatures leading to material aging and localized temperature rise causing thermal damage. Therefore, a thermal failure risk index can be defined as follows: ,in, The highest temperature, The temperature rise rate is obtained by differentiating the thermal field calculation results with respect to time. This refers to the duration of high temperature (within a safe range but with an operating temperature that is too high, such as above 100℃, which can be adjusted according to the actual situation). The function for calculating the thermal failure risk index is as follows: Define the highest temperature risk as ;in The maximum allowable operating temperature of the material. The ambient temperature.

[0082] Define the risk of temperature rise rate as , To allow the maximum rate of temperature rise.

[0083] Define the cumulative damage risk at high temperatures as ; Thermal aging damage can be calculated using the Arrhenius model. This represents the maximum permissible degree of thermal aging damage to a material during its design lifespan. It is typically set to 1 (indicating complete aging failure) or as specified in the material standard.

[0084] In a further implementation, the overall thermal risk is considered. , , and The weight coefficients for each indicator item satisfy the following conditions: .

[0085] 3.3. Based on the mechanical field calculation results, extract the maximum mechanical stress, structural deformation, and fatigue damage parameters to establish a mechanical failure risk assessment model. The mechanical risk index can then be defined as follows: ; Indicates the maximum mechanical stress. For the maximum displacement, This is fatigue damage. The function for calculating the mechanical failure risk index is as follows: Define stress risk index as , This represents the maximum allowable stress of the material.

[0086] Define the structural deformation risk index as , This represents the maximum allowable deformation of the material.

[0087] The risk of fatigue damage is calculated using a cumulative fatigue model: , For materials under mechanical stress The actual number of cycles under the influence, For corresponding stress The number of loops allowed at the horizontal level.

[0088] In a further implementation, comprehensive mechanical risks , , and The weight coefficients for each indicator item satisfy the following conditions: .

[0089] S4. In one implementation, a coupled risk objective is introduced, and a multi-objective optimization function is constructed using structural parameter variables as optimization variables; a dual-path solution strategy is constructed, including transforming the multi-objective optimization function into a single objective to obtain a single optimal solution, and obtaining the Pareto front of the multi-objective; According to S1, the set of optimizable parameter variables is defined as follows: Including the number of umbrella skirts , Apron diameter (outer diameter) Umbrella skirt spacing Sheath thickness mandrel dimensions (diameter) Equalizing ring dimensions (radius) Pipe diameter ) and installation location (installation height) Furthermore, considering the material parameters of the insulator... and heat dissipation structure parameters This constitutes the design variable space for insulators. .

[0090] 4.1 Based on the electrical failure risk, thermal failure risk, and mechanical failure risk calculated by S3, a multi-objective optimization function is constructed, expressed as: Specifically: For electrical reliability purposes, the goal is to reduce the maximum electric field strength, electric field gradient, and areas of electric field concentration. The thermal stability target aims to reduce maximum temperature, thermal aging loss, and temperature rise rate. The goal is to reduce maximum stress, structural deformation, and fatigue damage, in order to achieve mechanical reliability.

[0091] To achieve manufacturing cost targets, , , , These are material costs, processing costs, and heat dissipation structure costs, respectively.

[0092] In a further specific implementation, the manufacturing cost and the set of structural parameter variables... The correlation can be quantified in the following ways: Regarding material costs, ; The volume of the insulator (unit: cubic meters) is calculated from geometric parameters; the volume of each part (core rod, sheath, skirt, etc.) is calculated according to the volume function corresponding to the structural parameters.

[0093] Regarding processing costs, ; and These refer to the number of umbrella skirts and the number of equalizing rings, respectively. Regarding the cost of heat dissipation structure ; This represents the total structural length of the insulator.

[0094] , and These are the unit volume material cost coefficient (unit: yuan / cubic meter), the processing cost coefficient for a single umbrella skirt / equalizing ring, and the unit length design and assembly cost coefficient (unit: yuan / meter); all of the above cost coefficients are pre-determined based on actual production processes and market quotations.

[0095] Furthermore, to avoid the problem of failure caused by coupling effects despite low risk from a single physical field, a coupling risk objective is introduced. ; The electro-thermal-mechanical coupling influence factor is used to represent the synergistic risk generated when multiple failure factors increase simultaneously; the calculation method is as follows: .

[0096] 4.2 This implementation method employs two optional solution strategies, including the weighted summation method and the Pareto non-dominated sorting method. The specific choice depends on the actual application scenario.

[0097] Strategy 1: Applicable to engineering scenarios requiring a single optimal solution to be obtained quickly; it uses a weighted summation method to transform a multi-objective optimization problem into a single-objective optimization problem, i.e., expressed as... , For the first The normalized value of each objective item. For the first Dynamic weights of individual sub-objectives .

[0098] In a further implementation, considering the different sensitivities of failure modes under different operating conditions, the weight of each target item is dynamically adjusted according to the operating status.

[0099] First, establish the operating condition vector: establish the operating condition feature vector. , For voltage level, For ambient temperature, Indicates wind speed, For the level of filth, Indicates the level of mechanical load.

[0100] Furthermore, normalization is used to obtain dynamic weights; for the th There are several target items, each with a weight: ;in, For the first The original weight scores for each objective item are calculated as follows: ;in, For the first The feature is related to the first Sensitivity coefficient of each target item The first element in the working condition feature vector Nonlinear sensitivity functions with several characteristics; For the first The basic bias coefficient for each target item. The sensitivity coefficient and basic bias coefficient can be calibrated based on historical fault data. For example, for voltage characteristics under electrical target items, the influence coefficient can be set to 0.3; for load characteristics under mechanical target items, the sensitivity coefficient can be set to 0.5. The basic bias coefficient can be set according to the proportion of failure types in historical fault statistics.

[0101] In a further implementation, a nonlinear sensitivity function is used to map operating condition characteristics to a unified sensitivity metric space. In a specific implementation that may be referenced, the following typical sensitivity function is defined: (1) For operating voltage Define the sensitivity function ; (2) Regarding the level of filth Define the sensitivity function This simulates the pollution saturation effect; it should be noted that in actual function calculations, Quantification is achieved using equivalent salt density (ESDD) and gray density (NSDD). The base equivalent salt density is used to measure the degree of contamination on the outer insulation surface of the insulator. The salt density value corresponding to light contamination can be taken, such as 0.025 mg / cm2 or 0.05 mg / cm2.

[0102] (3) Regarding ambient temperature Define the sensitivity function This indicates that the risk increases dramatically under high-temperature operating conditions. This refers to the reference ambient temperature under normal design operating conditions for the insulator, such as 25℃ or 40℃. This is a temperature-sensitive scale parameter used to control the rate of increase in temperature sensitivity, and can be calibrated based on material thermal aging test data.

[0103] (4) For mechanical loads Define the sensitivity function Calibrate pseudo-nonlinear mechanical fatigue based on thermal aging test data of mold materials; The reference mechanical load indicates the rated mechanical failure load or rated tensile / bending load of the insulator, which is determined according to the insulator model.

[0104] (5) Regarding wind speed Define the sensitivity function ; The reference wind speed is used.

[0105] In one implementation description, during actual simulation, the above benchmark values ​​can be preset according to the power grid level, environmental meteorological data and material rated parameters of the specific application scenario of the insulator, and can be adjusted according to the set working conditions during implementation.

[0106] Strategy 2: It can be used in research scenarios that require a comprehensive consideration of multiple failure modes; it employs a multi-objective evolutionary algorithm to solve for the Pareto front solution set.

[0107] 4.3 Establish the optimization constraints that the optimization process needs to meet, including electric field constraints, temperature constraints, mechanical constraints, and manufacturing constraints; Electric field constraint is , Maximum allowable electric field strength; temperature constraint is , The maximum allowable operating temperature of the material; mechanical constraints are... and , and These are the maximum allowable stress and maximum allowable deformation of the material, respectively.

[0108] Manufacturing constraints represent the geometric boundary constraints that must be met for each part of the insulator during the actual manufacturing process. For example, the minimum wall thickness of the sheath must be no less than 3mm to prevent sheath breakage during the extrusion process; the ratio of the shed spacing to the diameter must be no less than 0.3 to ensure demolding capability; and the total length must be limited within the installation space to meet the size requirements of substations or transmission towers. These constraints are used as penalty terms in the simulation optimization process. When any manufacturing constraint is violated, a large penalty value is assigned to the objective function to eliminate the corresponding solution.

[0109] S5. In one implementation, a multi-objective surrogate model for solving optimization variables and risk indicators is established, and a dual-path iterative solution is performed. During the iteration process, the solution results of the dual paths are interactively verified, and the candidate solutions are fed back to the surrogate model for error calculation to update the training of the surrogate model until both paths converge and the optimal solution is output.

[0110] To address the issues of complex calculations and long computation times in multiphysics finite element models, a rapid optimization mechanism based on a surrogate model is established. This mechanism replaces numerous repetitive finite element solutions with finite sample learning, enabling rapid search for optimization solutions.

[0111] 5.1 Optimizing the variable space In this study, the Latin hypercube sampling method was used to generate the training sample set. ; For each group of samples ,calculate as well as This constitutes the training data.

[0112] 5.2 Establish a structural parameter-risk indicator proxy model, expressed as follows: ; Output for the proxy model, For surrogate prediction models, Gaussian process regression, neural networks, or support vector regression can be used. Output .

[0113] In a further implementation, for the same candidate solution, a single forward evaluation can obtain... Five types of objective values, corresponding to two solution strategies, will Supplying two parallel solution paths (single-objective, Pareto), sharing a surrogate model. Perform a simulation search; For path 1: This strategy path can use particle swarm optimization algorithms for fast search and output the optimal solution.

[0114] For path 2: This strategy path can be processed using the NSGA series of algorithms to generate a Pareto optimal solution set.

[0115] Each interval In the next iteration, path 1 outputs the current optimal solution. and its objective function value Path 2 outputs the current non-dominated solution set. .

[0116] 5.3. Perform interactive verification on the solutions output by both paths; The current optimal solution of path 1 Projected onto the Pareto front of path 2 In the process, determine whether it is controlled: like Not If any solution dominates, continue the current iteration; like quilt A certain solution exists in Dominate, trigger the current weight The adjustment mechanism.

[0117] Furthermore, the weight adjustment mechanism includes the following steps: Calculate the dominant solution Compared to Improvement rate of each target item : ; The weights are adjusted inversely based on the magnitude of improvement; for the first... Each objective item has an adjusted weight. , ; The learning rate is set to, for example, 0.1. The adjusted weights of each objective term are then normalized to ensure... .

[0118] Updated weights Feedback is sent to path 1 to restart single-objective optimization.

[0119] When the Pareto front region of path 2 is stable, i.e. continuous The rate of change of the frontier in the next iteration is less than the threshold. At that time, the TOPSIS method is used to extract solutions from the current Pareto set. A compromise solution is selected and injected into the initial population of path 1 to accelerate the convergence of the single-objective search of path 1.

[0120] In a specific implementation that can be referenced, two extreme solutions in the Pareto front (such as the solution with the minimum electrical risk index and the solution with the lowest cost) can also be selected as auxiliary initial points to increase the population diversity of path 1.

[0121] In a specific implementation that can be referenced, during the iteration process, the current optimal solution of path 1 and the solutions in the current non-dominated solution set of path 2 are used to construct a candidate point set. For each point in the candidate point set... Check its presence in the sample set Are there already real simulation results in China? If there are no simulation results, directly... Add to the set of points to be added ; If there are real simulation results Then calculate the predicted value of the proxy model. Compared with the true value relative error : ; To prevent control numbers with a denominator of 0; It is an L2 norm. If Greater than the error threshold Then Included in the set of points to be added Furthermore, regarding the set of added points... Remove duplicate points from the set Euclidean distance. One of the two points.

[0122] right Points in The multiphysics coupling simulation of S2 is invoked to calculate its true target output. Then the new sample points Add to training sample set The agent model was retrained based on the expanded training sample set.

[0123] When satisfied If the value is empty and no point addition is triggered for several consecutive rounds, point addition stops, and the optimization convergence check begins. If the single objective converges and the Pareto front converges, the final optimal solution / Pareto solution set is output, ending the entire optimization process. Otherwise, the process returns to constructing a candidate point set for iteration.

[0124] S6. In one implementation, the candidate solutions obtained by intelligent optimization are re-input into a high-precision multiphysics coupling model to verify electric field safety, temperature stability, mechanical reliability, and environmental adaptability.

[0125] 6.1 After convergence through step S5, path 1 outputs the optimal candidate solution. Path 2 outputs a compromise solution. The inputs are fed into the multiphysics coupling simulation model of S2, and the actual values ​​are output respectively. and .

[0126] At the same time, and Through proxy model Output the predicted values ​​respectively and ; right and The prediction error was calculated by calling the relative error calculation formula respectively. .

[0127] 6.2 From and Perform constraint checks as described in section 4.3 separately; when and / or of Less than or equal to the error threshold And if the constraint verification in section 4.3 is satisfied, then it is determined that... and / or The validation passed. The optimal solution is output. As a direct solution output. As a reference and aid in decision-making.

[0128] In one possible implementation, if and / or of Greater than the error threshold If any constraint is not met, a feedback correction mechanism is triggered. Specifically: If the constraint check is satisfied, but... Greater than the error threshold If so, it is considered that the surrogate model has insufficient local fitting. and / or Add to the training sample set for expanded training.

[0129] If the constraint verification is satisfied, but the calculated value of the risk indicator substituted into S3 deviates from the expected level beyond the expected value, then adjust the calculation parameters of the corresponding indicator in the S3 risk indicator calculation model and execute the verification in 6.1 again.

[0130] for ,like Less than or equal to the error threshold However, if multiple items (greater than or equal to 3 items) deviate from the constraint range to the set deviation, it can be determined that the weight allocation of multi-objective optimization is unreasonable. In this case, the weights are readjusted and feedback is sent to S5 to re-optimize.

[0131] As an embodiment of this application, a system for optimizing the design of ultra-high voltage insulators based on multi-field coupling is disclosed. The system, implemented using the specific embodiment described above, includes: The coupling space establishment module acquires the structural and material parameters of the composite insulator; defines the set of structural parameter variables and the set of material parameters, and establishes a dynamic correlation model between material parameters and temperature and field strength; and establishes multi-field data interaction relationships to model the multi-physics coupling parameter space. Multi-field coupled solution module; based on the coupled parameter space, a multi-field joint solution model is established; electric field calculation transmits the heat loss source to the thermal field, thermal field calculation feeds back material parameters to the electric field and transmits temperature load to the mechanical field; mechanical field calculation feeds back structural deformation information to the electric field and flow field, flow field calculation feeds back heat transfer boundary conditions to the thermal field; executes multi-field cyclic iteration and outputs multi-field data under steady state; A multi-objective optimization module is used to calculate risk indicators for electrical failure, thermal failure, and mechanical failure based on the multi-field data, introduce coupled risk objectives, and construct a multi-objective optimization function with structural parameter variables as optimization variables. A dual-path solution strategy is constructed, including transforming the multi-objective optimization function into a single objective to obtain a single optimal solution, and obtaining the Pareto front of the multi-objective. The optimization solution module establishes a multi-objective surrogate model for solving optimization variables and risk indicators, and performs dual-path iterative solutions. During the iteration process, the solution results of the dual paths are interactively verified, and candidate solutions are fed back to the surrogate model for error calculation to update the training of the surrogate model until both paths converge and the optimal solution is output.

[0132] As an embodiment of this application, an electronic device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the computer program is loaded onto the processor, it employs the specific implementation described in the above-described optimization method.

[0133] As an embodiment of this application, a computer-readable storage medium is provided, which stores a computer program, and when the computer program is executed by a processor, it employs the specific implementation described in the above-described optimization method.

[0134] 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 it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the protection scope of the claims of the present invention.

Claims

1. A design optimization method for ultra-high voltage insulators based on multi-field coupling, characterized in that, include: Obtain the structural and material parameters of the composite insulator; Define the set of structural parameter variables and the set of material parameters, and establish a dynamic correlation model between material parameters and temperature and field strength; establish multi-field data interaction relationships to model the multi-physics coupling parameter space; Based on the coupled parameter space, a multi-field joint solution model is established; the electric field calculation transfers the heat source of loss to the thermal field, and the thermal field calculation feeds back the material parameters to the electric field and transfers the temperature load to the mechanical field. Mechanical field calculations feed back structural deformation information to the electric and flow fields, while flow field calculations feed back heat transfer boundary conditions to the thermal field; multi-field iterative calculations are performed to output multi-field data under steady-state conditions. Based on the multi-field data, risk indicators for electrical failure, thermal failure, and mechanical failure are calculated respectively. Coupled risk objectives are introduced, and a multi-objective optimization function is constructed with structural parameter variables as optimization variables. A dual-path solution strategy is constructed, including transforming the multi-objective optimization function into a single objective to obtain a single optimal solution, and obtaining the Pareto front of the multi-objective. A multi-objective surrogate model for solving optimization variables and risk indicators is established, and a dual-path iterative solution is performed. During the iteration process, the solution results of the dual paths are interactively verified, and the candidate solutions are fed back to the surrogate model for error calculation to update the training of the surrogate model until both paths converge and the optimal solution is output.

2. The ultra-high voltage insulator design optimization method based on multi-field coupling according to claim 1, characterized in that, The process of defining the set of structural parameter variables includes: In the simulation analysis environment, the structural parameters of the composite insulator are constructed into a three-dimensional geometric model through the corresponding simulation mapping method; the three-dimensional geometric model is parameterized, and the optimizable structural parameters are defined as a set of structural parameter variables. Set corresponding physical constraint boundaries and coupling constraints for each variable in the set of structural parameter variables.

3. The ultra-high voltage insulator design optimization method based on multi-field coupling according to claim 1, characterized in that, The multiphysics coupling parameter space includes a set of structural parameter variables, a set of material parameters, a set of environmental parameters, and a dynamic correlation model of material parameters as a function of temperature and electric field. The dynamic correlation models of material parameters with temperature and electric field include temperature correlation models of relative permittivity, electrical conductivity, elastic modulus, and thermal expansion.

4. The ultra-high voltage insulator design optimization method based on multi-field coupling according to claim 1, characterized in that, The multi-field joint solution model is used to perform data interaction between different physical fields, including: An electric field model is established to calculate the electric field strength. The leakage current and the relative permittivity of the material are combined to calculate the leakage current heat source and the dielectric loss heat source, forming the total loss heat source transferred to the thermal field calculation. A thermal field model is established to calculate heat conduction based on the total heat loss source. Convection and radiation heat transfer boundaries are set, and the temperature distribution of the composite insulator is obtained by solving. The relative permittivity, conductivity and thermal expansion coefficient of the calculated material are updated based on the temperature distribution and fed back to the electric field model.

5. The ultra-high voltage insulator design optimization method based on multi-field coupling according to claim 4, characterized in that, The multi-field joint solution model is used to perform data interaction between different physical fields, and also includes: The thermal field model transfers the temperature distribution of the insulator to the mechanical field model; The mechanical field model calculates the thermal strain and structural displacement of the material based on the temperature distribution and the mechanical load during the operation of the insulator, updates the three-dimensional geometric model of the insulator, and transfers the thermal strain and structural displacement to the flow field model; A flow field model is established to calculate the velocity distribution and surface convection heat transfer coefficient based on the operating environment parameters, and then fed back to the thermal field model to update the heat transfer boundary conditions.

6. The ultra-high voltage insulator design optimization method based on multi-field coupling according to claim 1, characterized in that, The multi-field cyclic iteration includes: Based on the data interaction of the electric field model, thermal field model, mechanical field model, and flow field model, the comprehensive coupling error of each iteration is calculated; when the comprehensive coupling error is less than the convergence threshold, the multiphysics data of the current iteration is determined to be multiphysics data output in a steady state. The multi-field data under steady-state conditions include electric field distribution, temperature distribution, and stress distribution.

7. The ultra-high voltage insulator design optimization method based on multi-field coupling according to claim 1, characterized in that, The calculation process of the risk indicator includes: For electrical failure risk indicators, the maximum field strength, maximum field strength gradient, and area of ​​the concentrated electric field region are extracted from the electric field distribution in the multi-field data. For the thermal failure risk index, it is calculated based on the highest temperature, temperature rise rate and high temperature duration extracted from the temperature distribution in the multi-field data. For mechanical failure risk indicators, mechanical stress, maximum displacement, and fatigue damage are calculated based on the stress distribution in the data.

8. The ultra-high voltage insulator design optimization method based on multi-field coupling according to claim 1, characterized in that, The multi-objective optimization function includes: Minimize electrical failure risk indicators, thermal failure risk indicators, mechanical failure risk indicators, and manufacturing costs; it also includes coupled risk objectives, which aim to minimize coupled impact factors; The coupling influence factor is calculated based on the fusion of electrical failure risk indicators, thermal failure risk indicators, and mechanical failure risk indicators.

9. The ultra-high voltage insulator design optimization method based on multi-field coupling according to claim 1, characterized in that, The process of transforming a multi-objective optimization function into a single-objective solution for obtaining a single optimal solution includes: The electrical failure risk index, thermal failure risk index, mechanical failure risk index, manufacturing cost, and coupling influencing factors are weighted to obtain a single-objective optimization function; the weight of each objective item is dynamically adjusted according to the operating conditions.

10. The ultra-high voltage insulator design optimization method based on multi-field coupling according to claim 1, characterized in that, The interactive verification of the solution results for the two paths during the iteration process includes: Project the current optimal solution output by the single-objective path onto the Pareto front of the multi-objective path to determine the dominance relationship; If the current optimal solution is not dominated by any solution in the Pareto front, continue the current iteration; If the current optimal solution is dominated by at least one solution in the Pareto front, the improvement of the dominant solution relative to the current optimal solution in each objective term is calculated, and the weight of each objective term in the single objective path is adjusted according to the improvement.

11. A system for design optimization of ultra-high voltage insulators based on multi-field coupling, comprising executing the design optimization method as described in any one of claims 1-10, characterized in that, The system includes: The coupling space establishment module acquires the structural and material parameters of the composite insulator; defines the set of structural parameter variables and the set of material parameters, and establishes a dynamic correlation model between material parameters and temperature and field strength; and establishes multi-field data interaction relationships to model the multi-physics coupling parameter space. Multi-field coupled solution module; based on the coupled parameter space, a multi-field joint solution model is established; electric field calculation transmits the heat loss source to the thermal field, thermal field calculation feeds back material parameters to the electric field and transmits temperature load to the mechanical field; mechanical field calculation feeds back structural deformation information to the electric field and flow field, flow field calculation feeds back heat transfer boundary conditions to the thermal field; executes multi-field cyclic iteration and outputs multi-field data under steady state; A multi-objective optimization module is used to calculate risk indicators for electrical failure, thermal failure, and mechanical failure based on the multi-field data, introduce coupled risk objectives, and construct a multi-objective optimization function with structural parameter variables as optimization variables. A dual-path solution strategy is constructed, including transforming the multi-objective optimization function into a single objective to obtain a single optimal solution, and obtaining the Pareto front of the multi-objective. The optimization solution module establishes a multi-objective surrogate model for solving optimization variables and risk indicators, and performs dual-path iterative solutions. During the iteration process, the solution results of the dual paths are interactively verified, and candidate solutions are fed back to the surrogate model for error calculation to update the training of the surrogate model until both paths converge and the optimal solution is output.

12. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the computer program is loaded into the processor, it implements the design optimization method according to any one of claims 1-10.

13. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the design optimization method according to any one of claims 1-10.

Images