Structural parameter construction method for reducing vibration of transformer

Abstract

The invention relates to a structural parameter construction method for reducing vibration of a transformer. The method comprises the following steps: acquiring vibration characteristics and working condition data of the transformer; constructing a genetic algorithm model based on the objective function; candidate solution vibration performance verification based on finite element simulation; constructing a hybrid optimization strategy based on a genetic algorithm and finite element simulation; verifying and optimizing the scheme under a multi-dimensional working condition; outputting a vibration calculation result and evaluating a scheme. According to the structural parameter construction method for reducing the vibration of the transformer, the preliminary schemes are screened by applying a fast convergence genetic algorithm, and the candidate schemes are subjected to high-precision verification in combination with finite element simulation, so that the defects of a traditional vibration calculation method in the aspects of efficiency and accuracy are overcome; and the design change can be quickly responded, and the vibration performance is optimized, so that the high-standard requirements of modern power equipment on the aspects of safety and reliability are met, and powerful technical support is provided for the design and improvement of the transformer.

Description

Technical Field

[0001] The present invention relates to the technical field of transformers, and in particular to a method for constructing structural parameters for reducing transformer vibration. Background Art

[0002] In the power system, the transformer is an important power equipment, and its safety and stability are directly related to the normal operation of the entire system. During the operation of the transformer, under the action of complex electromagnetic fields and mechanical stress, different forms of vibration will be generated. These vibrations will not only affect the performance of the transformer, but also cause fatigue and damage to its internal structure, thereby affecting the life and safety of the equipment. Therefore, how to effectively calculate and analyze the vibration characteristics of the transformer has become an important topic of concern to researchers and engineering technicians. Traditional transformer vibration calculation methods usually rely on analytical methods or finite element simulation. Although these methods can describe the vibration characteristics in theory, they often have certain limitations in practical applications. Due to the complexity of the transformer structure and the variability of operating conditions, traditional vibration calculation methods are difficult to achieve a comprehensive evaluation under various working conditions without sacrificing calculation accuracy. At the same time, the traditional vibration calculation process usually relies on the experience of designers and a lot of manual debugging, which makes the entire design and evaluation process less efficient and difficult to quickly adapt to design requirements under different conditions.

[0003] Existing vibration calculation methods mainly include analytical method and finite element simulation method. The analytical method mainly calculates the vibration characteristics by establishing a simplified physical model and using relevant mathematical formulas and empirical models. The advantages of this method are a complete theoretical system and fast calculation speed, but its limitation is that the establishment of the model usually relies on a large number of assumptions and simplifications, and it is difficult to fully consider the coupling of multiple complex factors in the actual operating environment, resulting in the calculation results being limited in accuracy and stability. On the other hand, as a vibration calculation method with higher accuracy, the finite element simulation method can effectively describe the vibration behavior of complex structures. The finite element method decomposes the complex structure of the transformer into multiple finite elements, so that it can more accurately simulate the vibration characteristics under actual working conditions. However, due to the complex calculation process of finite element simulation, a large amount of computing resources are required when processing large-scale complex models, and there are also certain difficulties in parameter optimization. Therefore, although the finite element simulation method can provide higher calculation accuracy, its high time cost and computing resource consumption make it difficult to be widely adopted in actual engineering applications, especially in situations where multiple design parameters need to be quickly evaluated and optimized.

[0004] In the practical application of transformer vibration calculation, in addition to ensuring the calculation accuracy, the efficiency and applicability of the calculation must also be considered. With the continuous improvement of the design requirements of power equipment, traditional vibration calculation methods are difficult to meet the increasingly complex requirements in practical applications. For example, in the transformer design process, designers need to quickly find the best vibration control solution among multiple possible design solutions, which requires the vibration calculation method to not only accurately describe the vibration characteristics, but also complete the calculation within a reasonable time. However, the existing methods usually require multiple trial and error and repeated verification to obtain relatively satisfactory results, which not only consumes a lot of time, but also increases the complexity and cost of the design. In addition, when facing multi-variable and multi-objective design problems, traditional methods are often difficult to achieve global optimization and control, and are prone to fall into the dilemma of local optimal solutions, thereby affecting the stability and reliability of the design results. Therefore, how to improve the efficiency of vibration calculation while ensuring the calculation accuracy has become a key technical problem that needs to be solved in the field of transformer vibration calculation. In view of the above problems, there is an urgent need for a vibration calculation method that can take into account both calculation accuracy and efficiency to improve the overall performance and reliability of transformer design and evaluation.

[0005] For example, patent CN117852355A proposes a method for calculating the vibration response of transformer windings, which includes establishing a two-dimensional winding model, calculating the magnetic field, constructing a dynamic model, deriving dynamic equations, calculating rigidity parameters and equivalent stiffness matrix, obtaining correlation coefficients, calculating the distribution of alternating electric forces, and using it as an excitation to solve the vibration response. This method simplifies the traditional simulation analysis process and achieves fast and accurate vibration response solutions. However, the method proposed by patent CN117852355A often involves complex algorithm design to construct an accurate two-dimensional winding model and perform magnetic field calculations, and places high demands on computer performance. Such high-demand computing resources may significantly increase the overall implementation cost.

[0006] For example, patent CN106777812B proposes a method for calculating the short-circuit vibration characteristics of transformer windings, which includes using ANSYS to establish a two-dimensional finite element model and assign equivalent parameters, perform vibration characteristics analysis (including setting gravity acceleration and damping parameters, applying boundary conditions, transient analysis multi-load step loading solution), and use ANSYS post-processor to output vibration displacement, velocity and acceleration. This method simplifies the solution of winding vibration problems and realizes accurate dynamic analysis of short-circuit vibration characteristics. However, patent CN106777812B uses ANSYS software for finite element analysis. Although this method can accurately simulate the short-circuit vibration characteristics of the transformer, the purchase and use costs of ANSYS software are relatively high. In addition, the process of establishing and solving finite element models is usually complicated and requires a lot of time and human resources.

[0007] For example, patent CN116305879A proposes a transformer vibration calculation optimization method, which includes inputting transformer parameters, calculating short-circuit current and leakage magnetic field, establishing a dynamic model and calculating vibration displacement, while considering the separation of the coil and the pad. This method improves the accuracy of the analysis of winding vibration during a sudden short circuit of the transformer through precise calculation. However, in the actual application of patent CN116305879A, ensuring the accuracy of input transformer parameters and the accuracy of calculating leakage magnetic field may require detailed data preparation and powerful computing resources, which undoubtedly increases the difficulty and cost of this method in actual operation.

[0008] For example, patent CN116224171A proposes a transformer winding vibration calculation method, which involves obtaining the target current signal and coil position parameters, calculating the coil axial displacement to evaluate the total vibration response, and is used to measure the transformer's short-circuit resistance. This method improves the accuracy of the evaluation of the transformer winding vibration response through precise calculation. However, patent CN116224171A improves the accuracy of the evaluation of the transformer winding vibration response through precise calculation, and its effect depends on the precise measurement of the target current signal and coil position parameters. In actual operation, it may be challenging to accurately obtain these parameters, thereby limiting the effectiveness of the method.

[0009] For example, patent CN108984941A proposes a transformer winding vibration calculation method for transformer state analysis and winding vibration simulation in AC / DC mixed mode, which includes a calculation step that considers the influence of excitation state. This method can accurately evaluate the winding vibration displacement and provide a scientific basis for determining the structural stability of the transformer, with high scientificity and practicality. However, the method provided by patent CN108984941A is suitable for transformer state analysis and vibration simulation under AC / DC superposition voltage. In order to realize the analysis of this mode, special software tools and additional computing resources may be required. Summary of the invention

[0010] In view of the above-mentioned deficiencies in the prior art, the present invention provides a method for constructing structural parameters for reducing transformer vibration, which significantly reduces the consumption of computing resources and improves the accuracy and efficiency of vibration calculation.

[0011] The technical solution adopted by the present invention is:

[0012] A method for constructing structural parameters for reducing transformer vibration, comprising:

[0013] Step S1, obtaining transformer vibration characteristics and operating condition data;

[0014] Step S2, constructing a genetic algorithm model based on the objective function;

[0015] Step S3, verifying the vibration performance of the candidate solution based on finite element simulation;

[0016] Step S4, constructing a hybrid optimization strategy based on genetic algorithm and finite element simulation;

[0017] Step S5, scheme verification and optimization under multi-dimensional working conditions;

[0018] Step S6, vibration calculation result output and / or solution evaluation.

[0019] Preferably, step S1 obtains transformer vibration characteristics and operating data, the operating data including: measured data acquisition, historical design scheme data acquisition and simulation data acquisition;

[0020] Acquisition of measured data: sensors are used to collect vibration characteristics, stress distribution and electromagnetic characteristics data during transformer operation;

[0021] Acquisition of historical design scheme data, collection of historical transformer design schemes and their corresponding vibration calculation results; used for the selection of initial population and setting of objective function;

[0022] Simulation data acquisition: using the mathematical or physical model of the transformer established in the previous design or analysis process based on the actual size, winding structure, and core design of the transformer, the vibration characteristics and stress distribution simulation data are obtained through finite element simulation, and the simulation results are used as the input for genetic algorithm optimization; the vibration characteristics of different design schemes are predicted through simulation data, and compared with the measured data to verify the accuracy and stability of the simulation model.

[0023] Preferably, step S2, constructing a genetic algorithm model based on the objective function, includes:

[0024] Objective function definition, design the objective function according to the actual needs of transformer vibration calculation;

[0025] The objective function is usually used to characterize the performance index of the transformer vibration characteristics. The objective function is defined as minimizing the vibration frequency f and maximizing the structural stability S;

[0026] The objective function is used to quantify the vibration performance of the transformer and guide the design by setting a clear optimization direction. Specifically, minimizing the vibration frequency f can reduce the risk of resonance, thereby improving the smooth operation of the equipment; while maximizing the structural stability S enhances the reliability and fatigue resistance of the transformer in long-term operation.

[0027] Minimize the vibration frequency objective function:

[0028]

[0029] Variable explanation: x design variable vector;

[0030] nThe number of vibration modes;

[0031] ω i The weight factor of the i-th mode;

[0032] Δf i The frequency change of the i-th mode;

[0033] Maximize the structural stability objective function:

[0034]

[0035] Variable explanation: σ yield The yield level of the material;

[0036] σ max (x) Maximum stress value under a specific combination of design variables;

[0037] Comprehensive objective function:

[0038] F(x)=α 1 f(x)+α 2 ·S(x)-α 3 Δd(x)

[0039] Variable explanation: F(x) is a comprehensive objective function that combines the weighted sum of vibration frequency, structural stability, and displacement change; α 1 ,α 2 ,α 3 The weight factor in the objective function is set according to the actual design requirements; the displacement change Δd(x) is used to limit the vibration amplitude of the overall structure of the transformer;

[0040] Initial population setting: According to the measured data and historical design scheme data in the data acquisition phase, a suitable initial population is selected as the initial input of the genetic algorithm;

[0041] Based on the measured data and historical design scheme data in the data acquisition phase, by analyzing key indicators such as vibration characteristics, stress distribution and electromagnetic characteristics in these data, the schemes similar to the current design requirements are screened out as candidate schemes for the initial population;

[0042] The fitness value of each individual in the population is used to evaluate its compliance with the objective function;

[0043] The fitness evaluation formula is as follows:

[0044]

[0045] Variable explanation: Fitness(x) is the fitness measure of individual x in the population; F(x) is the value of the comprehensive objective function;

[0046] Genetic algorithm operation design: including the design of selection, crossover and mutation operations;

[0047] The selection operation is used to select high-quality individuals from the current population, the crossover operation is used to generate new individuals, and the mutation operation is used to introduce random disturbances to prevent falling into the local optimal solution;

[0048] The crossover operation generates a new individual by combining the features of two individuals, thereby expanding the search space;

[0049] The selection operation evaluates the fitness of individuals in the population, thereby screening out individuals with excellent performance;

[0050] The fitness value of each individual is calculated based on the degree of match between each individual and the objective function; the individuals are sorted by fitness value to ensure that individuals with high fitness have a greater probability of being retained or selected; based on the sorting, a certain probability strategy is used to determine which individuals enter the next generation, balancing the process of exploration and utilization, avoiding falling into the local optimal solution too early, and ensuring that the diversity of the population is preserved;

[0051] The crossover operation is used to generate new population individuals. The formula is as follows:

[0052] x new =β·x parent1 +(1-β)·x parent2

[0053] Variable explanation: x new Newly generated population individuals; x parent1 、x parent2 Two parent population individuals; β crossover factor, usually a random number between 0 and 1;

[0054] The mutation operation is used to introduce random perturbations to prevent the population from falling into the local optimal solution. The formula is as follows:

[0055] x mutated =x original +η·r

[0056] Variable explanation: x mutated The individual of the population after mutation; x original The individuals of the original population; η the variation amplitude factor; r the random number vector.

[0057] Preferably, step S3 is based on the finite element simulation candidate solution vibration performance verification, including:

[0058] Vibration performance simulation of candidate solutions: finite element simulation calculation is performed on the candidate solutions generated by the genetic algorithm to verify their vibration performance under actual working conditions, including vibration frequency and stress distribution;

[0059] The simulation results are compared with the measured data. The simulation results are compared and analyzed with the measured data to evaluate the accuracy and stability of the candidate solutions and determine whether the genetic algorithm model needs to be adjusted in parameters. In the process of comparing and analyzing the simulation results with the measured data, the performance indicators such as vibration frequency and stress distribution calculated by finite element simulation are compared with the actual measured data to evaluate the consistency between the two. If the deviation between the simulation results and the measured data is less than 10%, it indicates that the simulation model has high accuracy and stability and can reliably predict the vibration characteristics of the transformer. If the deviation is greater than 10%, adjust the transformer structure size of the simulation model.

[0060] Preferably, step S4 constructs a hybrid optimization strategy based on genetic algorithm and finite element simulation, including:

[0061] Genetic algorithm preliminary screening, using the global search capability of the genetic algorithm to conduct preliminary screening of the multidimensional parameter space and select candidate solutions that meet the requirements of the objective function; in the preliminary screening stage of the genetic algorithm, the fitness of the population individuals is evaluated through the objective function to screen out candidate solutions that meet the target requirements;

[0062] Define the comprehensive objective function F(x) to evaluate the fitness of individuals in the population;

[0063] F(x)=α 1 ·f vib (x)+α 2 ·S struct (x)-α 3 ·d max (x)

[0064] Variable interpretation: x is the design variable vector; f is the design variable vector; vib (x) The vibration frequency of the transformer under a specific combination of design variables; S struct (x) The stability of the transformer structure under a specific combination of design variables; d max (x) Maximum displacement of the transformer structure under a specific combination of design variables; α 1 ,α 2 ,α 3 The weight factor of the objective function;

[0065] Finite element simulation is used for precise verification. Finite element simulation analysis is performed on the candidate solutions initially selected. The performance of the candidate solutions under actual working conditions is verified through high-precision simulation calculations, and the optimal solution that meets the design requirements is selected.

[0066] Verify the performance of candidate solutions under actual working conditions through simulation calculations, conduct finite element analysis on selected candidate solutions to simulate the vibration frequency, stress distribution and displacement of the transformer under actual working conditions, and compare the simulation results with the measured data to ensure the accuracy and stability of the simulation model, so as to determine whether the candidate solutions meet the design requirements;

[0067] When performing finite element simulation on the candidate solutions initially screened by the genetic algorithm, the natural frequency of each candidate solution in a specific mode is calculated, and its structural stability and vibration characteristics are evaluated; the formula for defining the modal frequency is as follows:

[0068]

[0069] Variable explanation: f mode,i The natural frequency of the i-th mode; K i (x) Stiffness matrix of the structure under the i-th mode; M i (x) Mass matrix of the structure under the i-th mode;

[0070] Hybrid optimization strategy iteration: Continuously iterate the optimization strategy combining genetic algorithm and finite element simulation, adjust the genetic algorithm model parameters according to the simulation results, and improve the overall performance of vibration calculation;

[0071] The process of adjusting the genetic algorithm model parameters according to the simulation results is to optimize the fitness evaluation and operating parameters of the genetic algorithm based on the accuracy and stability verified by the simulation;

[0072] When the simulation results show that the performance of the candidate solution deviates from the objective function by more than 10%, the fitness function weight factor, selection rate, crossover rate or mutation rate in the genetic algorithm is adjusted;

[0073] In the iterative process of hybrid optimization strategy, the objective function and design variables are modified according to the error between the finite element simulation results and the output of the genetic algorithm model;

[0074] According to the error between the finite element simulation results and the output of the genetic algorithm model, the error is reduced by adjusting the weight factor of the objective function or correcting the value range of the design variable;

[0075] If the error is greater than 10%, increase the weight of the simulation verification result in the objective function to enhance its influence on the optimization direction; at the same time, fine-tune the design variables with large errors, narrow their search space or introduce local optimization operations;

[0076] The error correction formula is defined as follows:

[0077]

[0078] Variable explanation: x newThe design variable combination after error correction; x old The original population individuals generated by the genetic algorithm; η learning rate, used to control the error correction amplitude; E(x) error value, defined as the difference between the output of the genetic algorithm model and the finite element simulation result; The gradient of the error E(x) with respect to the design variable x is used to guide the individuals in the population to converge towards the global optimal solution.

[0079] Preferably, step S5 of scheme verification and optimization under multi-dimensional working conditions includes:

[0080] Multi-condition stability verification: verify the stability of the final output calculation scheme under different operating conditions to evaluate its adaptability to changes in the external environment; judge its stability by evaluating the response characteristics of the transformer under different operating conditions; define the following formula to calculate the vibration amplitude response under different conditions and evaluate its stability:

[0081]

[0082] Variable explanation: Δd response (C) Difference in vibration displacement response amplitude under specific working condition C, d i,max (C) The maximum vibration displacement value of the ith node under specific working condition C (obtained from finite element simulation or experimental data), d i,ref The maximum vibration displacement value of the i-th node under the reference working condition (usually the standard working condition or the rated working condition); N is the total number of vibration displacement nodes, C is the working condition parameter;

[0083] Computational complexity and resource usage evaluation: evaluate the complexity and resource usage of the computational solution, including computational time, computational resource requirements, and optimize the population size, crossover probability, and mutation probability of the genetic algorithm model. When evaluating the complexity and resource usage of the computational solution, the trade-off between computational time and computational resource requirements is used. Define the following formula to evaluate computational complexity and resource usage:

[0084] R comp =T calc ×C mem

[0085] Variable explanation: R comp Computational complexity and resource usage value; T calc Total time of genetic algorithm and simulation calculation (in seconds), including the time of fitness calculation, crossover, mutation and simulation verification for each individual in the population; C mem The maximum memory usage during the calculation;

[0086] Scheme optimization and output: further optimize the calculation scheme according to the verification results, and output the final scheme for practical application; when optimizing the calculation scheme, adjust the genetic algorithm model parameters according to the results of multi-condition stability verification and computational complexity evaluation, and define the following optimization objective function to guide the further optimization of the genetic algorithm:

[0087] min{α 1 ·Δd response (C)+α 2 ·R comp}

[0088] Variable explanation: Δd response (C) Difference in vibration displacement response under different working conditions C; R comp Computational complexity and resource usage; α 1 ,α 2 Weight factor used to balance the impact of vibration stability and computational complexity.

[0089] Preferably, step S6 of outputting vibration calculation results and evaluating solutions includes:

[0090] Output of vibration calculation results: Output the final vibration calculation results, including vibration frequency, stress distribution, structural stability analysis data, and form a detailed calculation report;

[0091] Scheme evaluation and optimization suggestions: Evaluate the scheme based on the calculation results and design goals, and put forward further optimization suggestions and improvement directions.

[0092] The beneficial effects of the present invention compared with the prior art are as follows:

[0093] The structural parameter construction method for reducing transformer vibration of the present invention provides a hybrid optimization strategy combining genetic algorithm and finite element simulation to achieve efficient and accurate calculation of transformer vibration. Preliminary solutions are screened by using a fast-converging genetic algorithm, and candidate solutions are verified with high precision in combination with finite element simulation, aiming to overcome the shortcomings of traditional vibration calculation methods in terms of efficiency and accuracy; it can quickly respond to design changes and optimize vibration performance, thereby meeting the high standards of modern power equipment in terms of safety and reliability, and providing strong technical support for the design and improvement of transformers.

[0094] The structural parameter construction method for reducing transformer vibration of the present invention utilizes the global search capability of genetic algorithms to effectively avoid the local optimal problems common in traditional methods in multi-dimensional parameter space, improves the global optimization capability of vibration calculation, and ensures the applicability and stability of the final solution under complex working conditions.

[0095] The present invention is based on a hybrid optimization strategy of genetic algorithm and finite element simulation, which combines the fast convergence characteristics of genetic algorithm with the high precision of finite element simulation, and adopts a screening-first-then-verification approach, which significantly reduces the consumption of computing resources and improves the accuracy and efficiency of vibration calculation.

[0096] The present invention is based on a comprehensive analysis method of vibration characteristics under multiple working conditions. It can comprehensively analyze and verify the vibration characteristics of the transformer under different operating conditions (such as load, temperature, etc.), and improve the applicability and reliability of the solution through the evaluation of multi-dimensional working conditions.

[0097] The present invention is based on the model calibration technology of measured data and simulation data, and proposes a model calibration method by combining measured data, historical design data and simulation data to achieve consistency between the genetic algorithm model and the actual working conditions, thereby ensuring the accuracy of the vibration calculation results.

[0098] The invention is based on an optimization control method for computing resources and time costs. By controlling parameters such as the population size, crossover probability and mutation probability of the genetic algorithm, the complexity of vibration calculation is optimized. Under the premise of ensuring the calculation accuracy, the calculation time and resource consumption are significantly reduced.

[0099] The present invention uses the global search capability of the genetic algorithm to quickly screen out candidate solutions that meet the requirements of the objective function, reducing the calculation time of multiple repeated experiments in the traditional finite element simulation method, and significantly improving the optimization efficiency.

[0100] The present invention combines genetic algorithm with finite element simulation and adopts a dual strategy of initial screening and precise verification, which effectively avoids the common model error accumulation problem in the prior art and ensures the accuracy and stability of vibration calculation results.

[0101] The present invention can perform vibration characteristic analysis and scheme verification under multi-dimensional working conditions (load, environment, frequency, etc.), thereby ensuring the applicability of the design scheme under complex working conditions.

[0102] The present invention significantly reduces the computing resource consumption of model optimization through the rapid convergence of the genetic algorithm. Compared with the high resource occupation of the finite element simulation method in the prior art, the present invention achieves more efficient computing resource allocation while meeting the accuracy requirements.

[0103] The hybrid optimization strategy of the present invention can quickly adapt to changes in different design parameters. Compared with the inefficient process of one-by-one verification in traditional methods, the present invention can quickly adjust the design scheme, shorten the design cycle, and improve the flexibility and reliability of the scheme. BRIEF DESCRIPTION OF THE DRAWINGS

[0104] Figure 1It is a flow chart of the transformer vibration calculation method based on the combination of genetic algorithm and finite element simulation. DETAILED DESCRIPTION

[0105] The present invention is described in detail below with reference to the accompanying drawings and embodiments:

[0106] Attached Figure 1 It can be seen that a method for constructing structural parameters for reducing transformer vibration includes:

[0107] Step S1, obtaining transformer vibration characteristics and operating condition data;

[0108] Step S2, constructing a genetic algorithm model based on the objective function;

[0109] Step S3, verifying the vibration performance of the candidate solution based on finite element simulation;

[0110] Step S4, constructing a hybrid optimization strategy based on genetic algorithm and finite element simulation;

[0111] Step S5, scheme verification and optimization under multi-dimensional working conditions;

[0112] Step S6: output of vibration calculation results and evaluation of solutions.

[0113] Preferably, step S1 of acquiring transformer vibration characteristics and operating condition data includes: acquiring measured data, acquiring historical design solution data, and acquiring simulation data;

[0114] Acquisition of measured data: Transformer operation is collected through sensors including vibration characteristics, stress distribution and electromagnetic characteristics data;

[0115] Acquisition of historical design scheme data, collection of historical transformer design schemes and their corresponding vibration calculation results; used for the selection of initial population and setting of objective function;

[0116] Simulation data acquisition: using the mathematical or physical model of the transformer established in the previous design or analysis process based on the actual size, winding structure, and core design of the transformer, the vibration characteristics and stress distribution simulation data are obtained through finite element simulation, and the simulation results are used as the input for genetic algorithm optimization; the vibration characteristics of different design schemes are predicted through simulation data, and compared with the measured data to verify the accuracy and stability of the simulation model.

[0117] The vibration characteristics directly reflect the dynamic response of the mechanical structure, are closely related to the noise and reliability of the transformer, and have a more direct impact on the final optimization goal. In contrast, although stress distribution and electromagnetic characteristics can provide more information, they are more used for the design optimization of local structures or electromagnetic performance analysis, and are not as directly and closely related to the main goal of vibration optimization as vibration characteristics. Therefore, simplifying the verification dimensions can improve efficiency and ensure clear focus.

[0118] Preferably, step S2, constructing a genetic algorithm model based on the objective function, includes:

[0119] Objective function definition, design the objective function according to the actual needs of transformer vibration calculation;

[0120] The objective functions include: minimizing vibration frequency and maximizing structural stability to ensure that the optimization direction of the genetic algorithm meets the transformer design objectives;

[0121] According to the actual needs of transformer vibration calculation, the objective function can be designed to minimize the vibration frequency of key components while ensuring structural stability; by setting a threshold or penalty function for the vibration frequency, the vibration is ensured to remain within a safe range, and structural stability constraints are added during the optimization process, such as limiting stress concentration or deformation. The optimization process can not only reduce the impact of vibration on the life of the equipment, but also ensure the stable operation of the transformer under different working conditions.

[0122] The objective function is usually used to characterize the performance index of the transformer vibration characteristics. The objective function is defined as minimizing the vibration frequency f and maximizing the structural stability S;

[0123] The objective function is used to quantify the vibration performance of the transformer and guide the design by setting clear optimization directions. Specifically, minimizing the vibration frequency (f) can reduce the risk of resonance and thus improve the smooth operation of the equipment. Maximizing the structural stability (S) enhances the reliability and fatigue resistance of the transformer in long-term operation.

[0124] This objective function ensures that the optimization process focuses on both vibration suppression and structural strength, thereby achieving a balance between performance and safety.

[0125] Minimize the vibration frequency objective function:

[0126]

[0127] Variable explanation: x design variable vector;

[0128] nThe number of vibration modes;

[0129] ω i The weight factor of the i-th mode;

[0130] Δf i The frequency change of the i-th mode;

[0131] Maximize the structural stability objective function:

[0132]

[0133] Variable explanation: σ yield The yield level of the material;

[0134] σ max (x) Maximum stress value under a specific combination of design variables;

[0135] Comprehensive objective function:

[0136] F(x)=α 1 f(x)+α 2 ·S(x)-α 3 Δd(x)

[0137] Variable explanation: F(x) is a comprehensive objective function that combines the weighted sum of vibration frequency, structural stability, and displacement change. 1 ,α 2 ,α 3 The weight factor in the objective function is set according to the actual design requirements; the displacement change Δd(x) is used to limit the vibration amplitude of the overall structure of the transformer;

[0138] Initial population setting: According to the measured data and historical design scheme data in the data acquisition phase, a suitable initial population is selected as the initial input of the genetic algorithm to improve the convergence speed and global search capability of the model;

[0139] Based on the measured data and historical design scheme data in the data acquisition phase, by analyzing key indicators such as vibration characteristics, stress distribution and electromagnetic characteristics in these data, the schemes similar to the current design requirements are screened out as candidate schemes for the initial population;

[0140] When selecting these solutions, priority is given to historical designs with better vibration performance or higher stability to ensure that these initial individuals can provide a good search starting point for the genetic algorithm, thereby accelerating the convergence of the model and improving the global search capability of the algorithm in complex design spaces.

[0141] The fitness value of each individual in the population is used to evaluate its compliance with the objective function;

[0142] The fitness evaluation formula is as follows:

[0143]

[0144] Variable explanation: Fitness(x) is the fitness measure of individual x in the population; F(x) is the value of the comprehensive objective function;

[0145] Genetic algorithm operation design: including the design of selection, crossover and mutation operations;

[0146] The selection operation is used to select high-quality individuals from the current population, the crossover operation is used to generate new individuals, and the mutation operation is used to introduce random disturbances to prevent falling into the local optimal solution. Through these operations, the population can be gradually optimized. The selection operation is used to select high-quality individuals because it can retain the solutions with better performance in the current population, ensuring that the optimization process moves in a better direction.

[0147] The crossover operation generates a new individual by combining the features of two individuals, thereby expanding the search space, promoting diversity and increasing the possibility of finding a better solution; while the mutation operation introduces random disturbances, which can prevent the algorithm from falling into the local optimum, help explore the global optimal solution, ensure the robustness of the algorithm and the quality of the final result, and approach the optimal solution of the objective function;

[0148] The selection operation evaluates the fitness of individuals in the population, thereby screening out individuals with excellent performance;

[0149] The fitness value of each individual is calculated based on the degree of match between each individual and the objective function; then, the individuals are sorted by fitness value to ensure that individuals with high fitness have a greater probability of being retained or selected; finally, based on the sorting, a certain probability strategy is used to determine which individuals enter the next generation, balancing the process of exploration and utilization, avoiding falling into the local optimal solution too early, and ensuring that the diversity of the population is preserved;

[0150] The crossover operation is used to generate new population individuals. The formula is as follows:

[0151] x new =β·x parent1 +(1-β)·x parent2

[0152] Variable explanation: x new Newly generated population individuals. parent1 、x parent2 Two parent population individuals; β crossover factor, usually a random number between 0 and 1;

[0153] The mutation operation is used to introduce random perturbations to prevent the population from falling into the local optimal solution. The formula is as follows:

[0154] x mutated =x original +η·r

[0155] Variable explanation: x mutated The individual of the population after mutation; xoriginal The individuals of the original population; η the variation amplitude factor; r the random number vector.

[0156] Preferably, step S3 is based on the finite element simulation candidate solution vibration performance verification, including:

[0157] Vibration performance simulation of candidate solutions: finite element simulation calculation is performed on the candidate solutions generated by the genetic algorithm to verify their vibration performance under actual working conditions, including vibration frequency and stress distribution;

[0158] Compare the simulation results with the measured data, compare and analyze the simulation results with the measured data, evaluate the accuracy and stability of the candidate solutions, and determine whether the genetic algorithm model needs to be adjusted. In the process of comparing and analyzing the simulation results with the measured data, compare the performance indicators such as vibration frequency and stress distribution calculated by finite element simulation with the actual measured data to evaluate the consistency between the two. If the deviation between the simulation results and the measured data is less than 10%, it indicates that the simulation model has high accuracy and stability and can reliably predict the vibration characteristics of the transformer. If the deviation is greater than 10%, adjust the transformer structure size of the simulation model to improve the simulation accuracy and make it closer to the actual working conditions.

[0159] Simulation feedback optimization returns the simulation verification results to the genetic algorithm model as the basis for adjusting the model parameters. By continuously optimizing the objective function and algorithm parameters, the performance of the genetic algorithm model under multi-dimensional working conditions is improved.

[0160] Preferably, step S4 constructs a hybrid optimization strategy based on genetic algorithm and finite element simulation, including:

[0161] Genetic algorithm preliminary screening, using the global search capability of the genetic algorithm to conduct preliminary screening of the multidimensional parameter space and select candidate solutions that meet the requirements of the objective function; in the preliminary screening stage of the genetic algorithm, the fitness of the population individuals is evaluated through the objective function to screen out candidate solutions that meet the target requirements;

[0162] Define the comprehensive objective function F(x) to evaluate the fitness of individuals in the population;

[0163] F(x)=α 1 ·f vib (x)+α 2 ·S struct (x)-α 3 ·d max (x)

[0164] Variable explanation: x is the design variable vector (such as transformer core size, winding position, conductor cross-sectional area, etc.); f is the design variable vector (such as transformer core size, winding position, conductor cross-sectional area, etc.); vib (x) The vibration frequency of the transformer under a specific combination of design variables; S struct(x) The stability of the transformer structure under a specific combination of design variables; d max (x) Maximum displacement of the transformer structure under a specific combination of design variables; α 1 ,α 2 ,α 3 The weight factor of the objective function;

[0165] Finite element simulation is used for precise verification. Finite element simulation analysis is performed on the candidate solutions initially selected. The performance of the candidate solutions under actual working conditions is verified through high-precision simulation calculations, and the optimal solution that meets the design requirements is selected.

[0166] Verify the performance of candidate solutions under actual working conditions through high-precision simulation calculations. Perform finite element analysis on the selected candidate solutions to simulate the vibration frequency, stress distribution and displacement of the transformer under actual working conditions. Compare the simulation results with the measured data to ensure the accuracy and stability of the simulation model, so as to determine whether the candidate solutions meet the design requirements.

[0167] When performing finite element simulation on the candidate solutions initially screened by the genetic algorithm, the natural frequency of each candidate solution in a specific mode is calculated, and its structural stability and vibration characteristics are evaluated; the formula for defining the modal frequency is as follows:

[0168]

[0169] Variable explanation: f mode,i The natural frequency of the i-th mode; K i (x) Stiffness matrix of the structure under the i-th mode; M i (x) Mass matrix of the structure under the i-th mode;

[0170] Hybrid optimization strategy iteration: Continuously iterate the optimization strategy combining genetic algorithm and finite element simulation, adjust the genetic algorithm model parameters according to the simulation results, and improve the overall performance of vibration calculation;

[0171] The process of adjusting the genetic algorithm model parameters according to the simulation results is mainly to optimize the fitness evaluation and operating parameters of the genetic algorithm based on the accuracy and stability verified by the simulation;

[0172] When the simulation results show that the performance of the candidate solution deviates from the objective function by more than 10%, adjust the fitness function weight factor, selection rate, crossover rate or mutation rate in the genetic algorithm to improve the ability to screen high-quality solutions and reduce the risk of falling into the local optimal solution, thereby improving the overall optimization effect of the model;

[0173] By increasing the weight factor related to the vibration frequency in the fitness function, the optimization process can be more focused on reducing the risk of resonance and improving the pertinence of the optimization results. In the selection operation, the selection rate is appropriately reduced to increase the diversity of the population, explore the solution space more deeply, and prevent premature convergence. Increasing the crossover rate helps to generate more new individuals, thereby expanding the search range and increasing the possibility of finding the global optimal solution. By increasing the mutation rate, more random disturbances can be introduced to effectively prevent the algorithm from falling into the local optimum. These adjustment strategies are crucial to ensuring the vibration calculation accuracy and stability of the transformer in the hybrid optimization process combining genetic algorithms and finite element simulation.

[0174] In the iterative process of hybrid optimization strategy, the objective function and design variables are modified according to the error between the finite element simulation results and the output of the genetic algorithm model;

[0175] According to the error between the finite element simulation results and the output of the genetic algorithm model, the error is reduced by adjusting the weight factor of the objective function or correcting the value range of the design variable, thereby improving the optimization accuracy;

[0176] If the error is greater than 10%, increase the weight of the simulation verification result in the objective function to enhance its influence on the optimization direction; at the same time, fine-tune the design variables with large errors, narrow their search space or introduce local optimization operations to better approach the actual optimal solution;

[0177] The error correction formula is defined as follows:

[0178]

[0179] Variable explanation: x new The design variable combination after error correction; x old The original population individuals generated by the genetic algorithm; η learning rate, used to control the error correction amplitude; E(x) error value, defined as the difference between the output of the genetic algorithm model and the finite element simulation result; The gradient of the error E(x) with respect to the design variable x is used to guide the individuals in the population to converge towards the global optimal solution.

[0180] Preferably, step S5 of scheme verification and optimization under multi-dimensional working conditions includes:

[0181] Multi-condition stability verification: verify the stability of the final output calculation scheme under different operating conditions to evaluate its adaptability to changes in the external environment; judge its stability by evaluating the response characteristics of the transformer under different operating conditions; define the following formula to calculate the vibration amplitude response under different conditions and evaluate its stability:

[0182]

[0183] Variable explanation: Δd response (C) Difference in vibration displacement response amplitude under specific working condition C, d i,max (C) The maximum vibration displacement value of the ith node under specific working condition C (obtained from finite element simulation or experimental data), d i,ref The maximum vibration displacement value of the i-th node under the reference working condition (usually the standard working condition or the rated working condition). N is the total number of vibration displacement nodes, C is the working condition parameter;

[0184] Computational complexity and resource occupancy evaluation: evaluate the complexity and resource occupancy of the computational solution, including computational time and computational resource requirements, and optimize the population size, crossover probability, and mutation probability of the genetic algorithm model to improve computational efficiency. When evaluating the complexity and resource occupancy of the computational solution, the trade-off between computational time and computational resource requirements is considered. Define the following formula to evaluate computational complexity and resource occupancy:

[0185] R comp =T calc ×C mem

[0186] Variable explanation: R comp Computational complexity and resource usage value; T calc Total time of genetic algorithm and simulation calculation (in seconds), including the time of fitness calculation, crossover, mutation and simulation verification for each individual in the population; C mem The maximum memory usage during the calculation;

[0187] Scheme optimization and output: further optimize the calculation scheme according to the verification results, and output the final scheme for practical application; when optimizing the calculation scheme, it is necessary to adjust the genetic algorithm model parameters according to the results of multi-condition stability verification and computational complexity evaluation, and define the following optimization objective function to guide the further optimization of the genetic algorithm:

[0188] min{α 1 ·Δd response (C)+α 2 ·R comp}

[0189] Variable explanation: Δd response (C) Difference in vibration displacement response under different working conditions C; R comp Computational complexity and resource usage; α 1 ,α 2 Weight factor used to balance the impact of vibration stability and computational complexity.

[0190] Preferably, step S6 of outputting the vibration calculation results and / or evaluating the solution includes:

[0191] Output of vibration calculation results: Output the final vibration calculation results, including vibration frequency, stress distribution, structural stability analysis data, and form a detailed calculation report;

[0192] Scheme evaluation and optimization suggestions: input the calculation results into the pre-built design target parameter evaluation model, compare the output vibration frequency, stress distribution, and structural stability analysis data with the data in the evaluation model one by one, and based on the comparison results, determine the calculation results and design goals to evaluate the scheme, put forward further optimization suggestions and improvement directions, and provide guidance for subsequent design.

[0193] The structural parameter construction method for reducing transformer vibration of the present invention utilizes the global search capability of genetic algorithms to effectively avoid the local optimal problems common in traditional methods in multi-dimensional parameter space, improves the global optimization capability of vibration calculation, and ensures the applicability and stability of the final solution under complex working conditions.

[0194] The present invention is based on a hybrid optimization strategy of genetic algorithm and finite element simulation, which combines the fast convergence characteristics of genetic algorithm with the high precision of finite element simulation, and adopts a screening-first-then-verification approach, which significantly reduces the consumption of computing resources and improves the accuracy and efficiency of vibration calculation.

[0195] The present invention is based on a comprehensive analysis method of vibration characteristics under multiple working conditions. It can comprehensively analyze and verify the vibration characteristics of the transformer under different operating conditions (such as load, temperature, etc.), and improve the applicability and reliability of the solution through the evaluation of multi-dimensional working conditions.

[0196] The present invention is based on the model calibration technology of measured data and simulation data, and proposes a model calibration method by combining measured data, historical design data and simulation data to achieve consistency between the genetic algorithm model and the actual working conditions, thereby ensuring the accuracy of the vibration calculation results.

[0197] The invention is based on an optimization control method for computing resources and time costs. By controlling parameters such as the population size, crossover probability and mutation probability of the genetic algorithm, the complexity of vibration calculation is optimized. Under the premise of ensuring the calculation accuracy, the calculation time and resource consumption are significantly reduced.

[0198] The present invention uses the global search capability of the genetic algorithm to quickly screen out candidate solutions that meet the requirements of the objective function, reducing the calculation time of multiple repeated experiments in the traditional finite element simulation method, and significantly improving the optimization efficiency.

[0199] The present invention combines genetic algorithm with finite element simulation and adopts a dual strategy of initial screening and precise verification, which effectively avoids the common model error accumulation problem in the prior art and ensures the accuracy and stability of vibration calculation results.

[0200] The present invention can perform vibration characteristic analysis and scheme verification under multi-dimensional working conditions (load, environment, frequency, etc.), thereby ensuring the applicability of the design scheme under complex working conditions.

[0201] The present invention significantly reduces the computing resource consumption of model optimization through the rapid convergence of the genetic algorithm. Compared with the high resource occupation of the finite element simulation method in the prior art, the present invention achieves more efficient computing resource allocation while meeting the accuracy requirements.

[0202] The hybrid optimization strategy of the present invention can quickly adapt to changes in different design parameters. Compared with the inefficient process of one-by-one verification in traditional methods, the present invention can quickly adjust the design scheme, shorten the design cycle, and improve the flexibility and reliability of the scheme.

[0203] The structural parameter construction method for reducing transformer vibration according to the present invention can be understood as a dynamic combination of some implementation details in some embodiments, especially when it comes to relevant algorithms, calculation methods, or software systems or devices. The key variables in the system or method can be set, adjusted or optimized according to the details of the construction method, so as to achieve those expected technical effects. Some details of the parameter construction method of the present invention can be understood as parameters in a model that can be defined in some embodiments, especially in those embodiments such as machine learning, the details of the present invention can be used to describe how to select and adjust the hyperparameters of a neural network. In some embodiments of the present invention, those details can be defined as key parameters in an algorithm, or configuration parameters between various components of a system, and the overall performance of the system or method can be optimized by these parameters. In some embodiments of the present invention, the method steps or details of the present invention or their permutations and combinations can be described by modules or devices, such as the technical details or features of the nth step implemented by the nth device or the nth module. In certain method embodiments of the method of the present invention, the result of the final step of the arrangement and combination can be output to the manufacturing device or equipment of the corresponding transformer or transformer component to achieve automatic production, or the result can be fed back to the initial step at the same time to further form a more optimal structural parameter construction, and after multiple iterations, the result of the final step can be output to the manufacturing device or equipment of the corresponding transformer or transformer component to achieve more optimal automatic production.

[0204] The above is only a preferred embodiment of the present invention, and does not limit the structure of the present invention in any form. Any simple modification, equivalent change and modification made to the above embodiment according to the technical essence of the present invention are within the scope of the technical solution of the present invention.

Claims

1. A method for constructing structural parameters to reduce transformer vibration, characterized in that: include: Step S1, obtaining transformer vibration characteristics and operating condition data; Step S2, constructing a genetic algorithm model based on the objective function; Step S3, verifying the vibration performance of the candidate solution based on finite element simulation; Step S4, constructing a hybrid optimization strategy based on genetic algorithm and finite element simulation; Step S5, scheme verification and optimization under multi-dimensional working conditions; Step S6, vibration calculation result output and / or solution evaluation.

2. The method for constructing structural parameters for reducing transformer vibration according to claim 1, characterized in that: Step S1: obtaining transformer vibration characteristics and operating data, the operating data including: actual measurement data, historical design data and simulation data; Acquisition of measured data: sensors are used to collect vibration characteristics, stress distribution and electromagnetic characteristics data during transformer operation; Acquisition of historical design scheme data, collection of historical transformer design schemes and their corresponding vibration calculation results; used for the selection of initial population and setting of objective function; Simulation data acquisition: using the mathematical or physical model of the transformer established in the previous design or analysis process based on the actual size, winding structure, and core design of the transformer, the vibration characteristics and stress distribution simulation data are obtained through finite element simulation, and the simulation results are used as the input for genetic algorithm optimization; the vibration characteristics of different design schemes are predicted through simulation data, and compared with the measured data to verify the accuracy and stability of the simulation model.

3. The method for constructing structural parameters for reducing transformer vibration according to claim 1, characterized in that: Step S2, constructing a genetic algorithm model based on the objective function, including: Objective function definition, design the objective function according to the actual needs of transformer vibration calculation; The objective function is usually used to characterize the performance index of the transformer vibration characteristics. The objective function is defined as minimizing the vibration frequency f and maximizing the structural stability S; The objective function is used to quantify the vibration performance of the transformer and guide the design by setting a clear optimization direction. Specifically, minimizing the vibration frequency f can reduce the risk of resonance, thereby improving the smooth operation of the equipment; while maximizing the structural stability S enhances the reliability and fatigue resistance of the transformer in long-term operation. Minimize the vibration frequency objective function: Variable explanation: x design variable vector; nThe number of vibration modes; ω i The weight factor of the i-th mode; Δf i The frequency change of the i-th mode; Maximize the structural stability objective function: Variable explanation: σ yield The yield level of the material; σ max (x) Maximum stress value under a specific combination of design variables; Comprehensive objective function: F(x)=α1·f(x)+α2·S(x)-α3·Δd(x) Variable explanation: F(x) is a comprehensive objective function, which is a weighted sum of vibration frequency, structural stability and displacement change; the weight factors in the α1, α2, and α3 objective functions are set according to actual design requirements; Δd(x) is the displacement change, which is used to limit the vibration amplitude of the overall transformer structure; Initial population setting: According to the measured data and historical design scheme data in the data acquisition phase, a suitable initial population is selected as the initial input of the genetic algorithm; Based on the measured data and historical design scheme data in the data acquisition phase, by analyzing key indicators such as vibration characteristics, stress distribution and electromagnetic characteristics in these data, the schemes similar to the current design requirements are screened out as candidate schemes for the initial population; The fitness value of each individual in the population is used to evaluate its compliance with the objective function; The fitness evaluation formula is as follows: Variable explanation: Fitness(x) is the fitness measure of individual x in the population; F(x) is the value of the comprehensive objective function; Genetic algorithm operation design: including the design of selection, crossover and mutation operations; The selection operation is used to select high-quality individuals from the current population, the crossover operation is used to generate new individuals, and the mutation operation is used to introduce random disturbances to prevent falling into the local optimal solution; The crossover operation generates a new individual by combining the features of two individuals, thereby expanding the search space; The selection operation evaluates the fitness of individuals in the population, thereby screening out individuals with excellent performance; The fitness value of each individual is calculated based on the degree of match between each individual and the objective function; the individuals are sorted by fitness value to ensure that individuals with high fitness have a greater probability of being retained or selected; based on the sorting, a certain probability strategy is used to determine which individuals enter the next generation, balancing the process of exploration and utilization, avoiding falling into the local optimal solution too early, and ensuring that the diversity of the population is preserved; The crossover operation is used to generate new population individuals. The formula is as follows: x new =β·x parent1 +(1-β)·x parent2 Variable explanation: x new Newly generated population individuals; x parent1 、x parent2 Two parent population individuals; β crossover factor, usually a random number between 0 and 1; The mutation operation is used to introduce random perturbations to prevent the population from falling into the local optimal solution. The formula is as follows: x mutated =x original +η·r Variable explanation: x mutated The individual of the population after mutation; x original The individuals of the original population; η the variation amplitude factor; r the random number vector.

4. The method for constructing structural parameters for reducing transformer vibration according to claim 1, characterized in that: Step S3: Verification of vibration performance of candidate solutions based on finite element simulation, including: Vibration performance simulation of candidate solutions: finite element simulation calculation is performed on the candidate solutions generated by the genetic algorithm to verify their vibration performance under actual working conditions, including vibration frequency and stress distribution; The simulation results are compared with the measured data. The simulation results are compared and analyzed with the measured data to evaluate the accuracy and stability of the candidate solutions and determine whether the genetic algorithm model needs to be adjusted in parameters. In the process of comparing and analyzing the simulation results with the measured data, the performance indicators such as vibration frequency and stress distribution calculated by finite element simulation are compared with the actual measured data to evaluate the consistency between the two. If the deviation between the simulation results and the measured data is less than 10%, it indicates that the simulation model has high accuracy and stability and can reliably predict the vibration characteristics of the transformer. If the deviation is greater than 10%, adjust the transformer structure size of the simulation model.

5. The method for constructing structural parameters for reducing transformer vibration according to claim 1, characterized in that: Step S4 constructs a hybrid optimization strategy based on genetic algorithm and finite element simulation, including: Genetic algorithm preliminary screening, using the global search capability of the genetic algorithm to conduct preliminary screening of the multidimensional parameter space and select candidate solutions that meet the requirements of the objective function; in the preliminary screening stage of the genetic algorithm, the fitness of the population individuals is evaluated through the objective function to screen out candidate solutions that meet the target requirements; Define the comprehensive objective function F(x) to evaluate the fitness of individuals in the population; F(x)=α1·f vib (x)+α2·S struct (x)-α3·d max (x) Variable interpretation: x is the design variable vector; f is the design variable vector; vib (x) The vibration frequency of the transformer under a specific combination of design variables; S struct (x) The stability of the transformer structure under a specific combination of design variables; d max (x) Maximum displacement of the transformer structure under a specific combination of design variables; α1, α2, α3 weight factors of the objective function; Finite element simulation is used for precise verification. Finite element simulation analysis is performed on the candidate solutions initially selected. The performance of the candidate solutions under actual working conditions is verified through high-precision simulation calculations, and the optimal solution that meets the design requirements is selected. Verify the performance of candidate solutions under actual working conditions through simulation calculations, conduct finite element analysis on selected candidate solutions to simulate the vibration frequency, stress distribution and displacement of the transformer under actual working conditions, and compare the simulation results with the measured data to ensure the accuracy and stability of the simulation model, so as to determine whether the candidate solutions meet the design requirements; When performing finite element simulation on the candidate solutions initially screened by the genetic algorithm, the natural frequency of each candidate solution in a specific mode is calculated, and its structural stability and vibration characteristics are evaluated; the formula for defining the modal frequency is as follows: Variable explanation: f mode,i The natural frequency of the i-th mode; K i (x) Stiffness matrix of the structure under the i-th mode; M i (x) Mass matrix of the structure under the i-th mode; Hybrid optimization strategy iteration: Continuously iterate the optimization strategy combining genetic algorithm and finite element simulation, adjust the genetic algorithm model parameters according to the simulation results, and improve the overall performance of vibration calculation; The process of adjusting the genetic algorithm model parameters according to the simulation results is to optimize the fitness evaluation and operating parameters of the genetic algorithm based on the accuracy and stability verified by the simulation; When the simulation results show that the performance of the candidate solution deviates from the objective function by more than 10%, the fitness function weight factor, selection rate, crossover rate or mutation rate in the genetic algorithm is adjusted; In the iterative process of hybrid optimization strategy, the objective function and design variables are modified according to the error between the finite element simulation results and the output of the genetic algorithm model; According to the error between the finite element simulation results and the output of the genetic algorithm model, the error is reduced by adjusting the weight factor of the objective function or correcting the value range of the design variable; If the error is greater than 10%, increase the weight of the simulation verification result in the objective function to enhance its influence on the optimization direction; at the same time, fine-tune the design variables with large errors, narrow their search space or introduce local optimization operations; The error correction formula is defined as follows: Variable explanation: x new The design variable combination after error correction; x old The original population individuals generated by the genetic algorithm; η learning rate, used to control the error correction amplitude; E(x) error value, defined as the difference between the output of the genetic algorithm model and the finite element simulation result; The gradient of the error E(x) with respect to the design variable x is used to guide the individuals in the population to converge towards the global optimal solution.

6. The method for constructing structural parameters for reducing transformer vibration according to claim 1, characterized in that: Step S5: Scheme verification and optimization under multi-dimensional working conditions, including: Multi-condition stability verification: verify the stability of the final output calculation scheme under different operating conditions to evaluate its adaptability to changes in the external environment; judge its stability by evaluating the response characteristics of the transformer under different operating conditions; define the following formula to calculate the vibration amplitude response under different conditions and evaluate its stability: Variable explanation: Δd response (C) Difference in vibration displacement response amplitude under specific working condition C, d i,max (C) The maximum vibration displacement value of the ith node under specific working condition C (obtained from finite element simulation or experimental data), d i,ref The maximum vibration displacement value of the i-th node under the reference working condition (usually the standard working condition or the rated working condition); N is the total number of vibration displacement nodes, C is the working condition parameter; Computational complexity and resource usage evaluation: evaluate the complexity and resource usage of the computational solution, including computational time, computational resource requirements, and optimize the population size, crossover probability, and mutation probability of the genetic algorithm model. When evaluating the complexity and resource usage of the computational solution, the trade-off between computational time and computational resource requirements is used. Define the following formula to evaluate computational complexity and resource usage: R comp =T calc ×C mem Variable explanation: R comp Computational complexity and resource usage value; T calc Total time of genetic algorithm and simulation calculation (in seconds), including the time of fitness calculation, crossover, mutation and simulation verification for each individual in the population; C mem The maximum memory usage during the calculation; Scheme optimization and output: further optimize the calculation scheme according to the verification results, and output the final scheme for practical application; when optimizing the calculation scheme, adjust the genetic algorithm model parameters according to the results of multi-condition stability verification and computational complexity evaluation, and define the following optimization objective function to guide the further optimization of the genetic algorithm: min{α1·Δd response (C)+α2·R comp } Variable explanation: Δd response (C) Difference in vibration displacement response under different working conditions C; R comp Computational complexity and resource occupancy value; α1, α2 weight factors are used to weigh the impact of vibration stability and computational complexity.

7. The method for constructing structural parameters for reducing transformer vibration according to claim 1, characterized in that: Step S6: Output of vibration calculation results and evaluation of solutions, including: Output of vibration calculation results: Output the final vibration calculation results, including vibration frequency, stress distribution, structural stability analysis data, and form a detailed calculation report; Scheme evaluation and optimization suggestions: Evaluate the scheme based on the calculation results and design goals, and put forward further optimization suggestions and improvement directions.

Images