A method, system, device and medium for optimizing the size of transformer reinforcing ribs

Abstract

The present invention belongs to the technical field of explosion-proof of power system transformers, and discloses a method, system, device and medium for optimizing the size of transformer stiffeners. The optimization method includes: determining the parameters to be optimized and the design space of the transformer stiffeners; the parameters to be optimized of the transformer stiffeners include the length of the stiffeners, the height of the stiffeners and the wall thickness of the stiffeners; collecting samples in the design space to obtain sampling samples; based on the sampling samples, through a fluid-structure interaction model, performing a simulation calculation on the structural deformation of the transformer tank wall to obtain the deformation amount of the transformer oil tank wall; fitting the non-linear function relationship between the parameters to be optimized of the transformer stiffeners and the deformation amount of the transformer oil tank wall to obtain a surrogate function; based on the surrogate function, performing parameter optimization to determine the size of the transformer stiffeners. The present invention significantly improves the explosion resistance of the transformer oil tank and reduces the risk of combustion and explosion caused by internal arc faults on the basis of effectively reducing the calculation cost and resource consumption.

Description

Technical Field

[0001] The present invention belongs to the technical field of explosion-proof technology for power system transformers, and relates to a method, system, device and medium for optimizing the size of transformer stiffeners. Background Art

[0002] Transformers are key equipment widely used in the power system and play important roles such as voltage level conversion and electric energy distribution. In recent years, fire and explosion accidents caused by internal oil arc faults in transformers have occurred frequently, seriously affecting the safe and stable operation of the power grid and causing huge economic losses. The insufficient explosion-proof ability of the transformer tank structure is an important reason for the deformation and rupture of the transformer box body, which further evolves into fire and explosion accidents. Therefore, improving the explosion-proof ability of the transformer tank structure has become a key issue that urgently needs to be solved to prevent transformer explosion accidents and improve the safe service level of large oil-filled equipment.

[0003] Currently, there is a lack of an effective guiding method for the explosion-proof design of the transformer tank structure. Most studies mainly rely on the engineering experience of operators, thickening the wall of the transformer tank cavity or adding stiffeners to enhance the structural strength of the transformer tank wall, but cannot reduce the risk of transformer explosion accidents. Summary of the Invention

[0004] Aiming at the problem that the design of the transformer tank wall stiffeners depends on manual experience setting and cannot reduce the risk of transformer explosion accidents, the purpose of the present invention is to provide a method, system, device and medium for optimizing the size of transformer stiffeners. This method can improve the anti-explosion ability of the mechanical structure of the transformer tank, reduce the deformation of the tank wall under internal arc faults, thereby reducing the risk of transformer explosion accidents. At the same time, it can also provide a theoretical basis for the design and optimization of the transformer box structure, which is of great significance and application value for improving the safe service level of newly built and in-service transformers and avoiding secondary explosion accidents.

[0005] To achieve the above purpose, the technical solution of the present invention is as follows:

[0006] The first aspect of the present invention provides a method for optimizing the size of transformer stiffeners, including the following steps:

[0007] Determine the parameters to be optimized and the design space of the transformer stiffeners; among them, the parameters to be optimized of the transformer stiffeners include the length of the stiffeners, the height of the stiffeners, and the wall thickness of the stiffeners;

[0008] Collect samples in the design space to obtain sampling samples;

[0009] Based on the sampling samples, through the fluid-structure interaction model, perform simulation calculations on the deformation of the transformer tank wall structure to obtain the deformation amount of the transformer tank wall;

[0010] Fit the non - linear function relationship between the parameters to be optimized of the transformer stiffener and the deformation of the transformer oil tank wall to obtain a surrogate function;

[0011] Based on the surrogate function, conduct parameter optimization to obtain the numerical values of the parameters to be optimized and determine the dimensions of the transformer stiffener.

[0012] Furthermore, the obtaining of sampling samples by sampling in the design space is carried out by using the Monte Carlo sampling method.

[0013] Furthermore, based on the sampling samples, through the fluid - solid coupling model, conduct the simulation calculation of the transformer tank wall structure deformation to obtain the deformation of the transformer oil tank wall, including the following steps:

[0014] Based on a sampling sample, build a three - dimensional geometric model of the transformer oil tank, including a fluid domain model and a solid domain model;

[0015] Solve the fluid domain model after updating the fluid domain boundary to obtain the fluid domain pressure distribution results and displacement data at the current moment;

[0016] Set the material properties of the transformer oil tank solid structure and the contact surface between the fluid domain and the solid domain;

[0017] Based on the fluid domain pressure distribution results and displacement data at the current moment, solve the solid domain model to obtain the deformation of the transformer oil tank wall at the current moment;

[0018] Use the dynamic mesh method to update the fluid domain boundary in the fluid domain model at the next moment, complete the transfer of the pressure distribution results and displacement data between the fluid domain and the solid domain, and obtain the deformation of the transformer oil tank wall at the final moment;

[0019] Conduct the calculation of the next sampling sample until the calculation of all sampling samples is completed to obtain the deformations of the transformer oil tank walls of all sampling samples.

[0020] Furthermore, the fitting of the non - linear function relationship between the parameters to be optimized of the transformer stiffener and the deformation of the transformer oil tank wall to obtain a surrogate function includes the following steps:

[0021] Form a surrogate function data set with each sampling sample and the deformation of the transformer oil tank wall corresponding to each sampling sample;

[0022] According to the surrogate function data set, construct a surrogate function through the least - squares support vector regression model.

[0023] Furthermore, the surrogate function is:

[0024]

[0025] Where, is a surrogate function, is the input dataset of the surrogate function, is the weight vector of the complexity of the least squares support vector regression model, w T is the transposed vector, is the characteristic function corresponding to the Sigmoid kernel function, is the function bias term, n is the total number of input data samples, is the Lagrange multiplier, is the radial basis function, is the th data sample in the input dataset of the surrogate function, is the th data sample in the input dataset of the surrogate function, is the serial number of the first data sample, is the serial number of the second data sample, is different from .

[0026] Furthermore, the parameter optimization based on the surrogate function to obtain the numerical values of the parameters to be optimized is performed using the particle swarm optimization algorithm.

[0027] Furthermore, the goal of parameter optimization based on the surrogate function is to minimize the deformation of the transformer tank wall.

[0028] The second aspect of the present invention provides a transformer stiffener size optimization system, including:

[0029] A module for determining parameters to be optimized and design space, which is used to determine the parameters to be optimized and the design space of the transformer stiffener; among them, the parameters to be optimized of the transformer stiffener include the length of the stiffener, the height of the stiffener, and the wall thickness of the stiffener;

[0030] A sample collection module, which is used to collect samples in the design space to obtain sampling samples;

[0031] A simulation calculation module, which is used to perform simulation calculations on the deformation of the transformer tank wall structure based on the sampling samples through a fluid-structure interaction model to obtain the deformation amount of the transformer tank wall;

[0032] A fitting module, which is used to fit the non-linear function relationship between the parameters to be optimized of the transformer stiffener and the deformation amount of the transformer tank wall to obtain a surrogate function;

[0033] A parameter optimization module, which is used to perform parameter optimization based on the surrogate function to obtain the numerical values of the parameters to be optimized and determine the size of the transformer stiffener.

[0034] A third aspect of the present invention provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the transformer reinforcing rib size optimization method is implemented.

[0035] A fourth aspect of the present invention provides a computer-readable storage medium storing a computer program, and when the computer program is executed by a processor, the transformer reinforcing rib size optimization method is implemented.

[0036] Compared with the prior art, the present invention has the following beneficial effects:

[0037] In the present invention, based on sampling samples, through a fluid-structure interaction model, the deformation simulation calculation of the transformer tank wall structure is carried out to obtain the deformation amount of the transformer oil tank wall; the non-linear function relationship between the parameters to be optimized of the transformer reinforcing rib and the deformation amount of the transformer oil tank wall is fitted to obtain a surrogate function; based on the surrogate function, parameter optimization is carried out to obtain the numerical values of the parameters to be optimized, realizing the optimization of the transformer reinforcing rib size. Different from the current prior art that relies on engineering experience and intuition to thicken the oil tank wall, the present invention innovatively adopts a fluid-structure interaction model that takes into account the dynamic behavior of the oil pressure inside the oil tank and the mechanical response of the tank body structure. Through simulation calculation, the influence law of different reinforcing rib size parameters on the plastic deformation of the oil tank wall is accurately quantified, so as to determine the transformer reinforcing rib size, providing a scientific theoretical guidance for the design of the transformer oil tank wall reinforcing rib. Based on the surrogate function, the present invention can greatly reduce the calculation cost and save computing resources, providing a reliable anti-explosion performance optimization design scheme for large oil-filled equipment.

[0038] Furthermore, the present invention takes the minimization of the deformation of the oil tank cavity wall as the goal, carries out optimization design for the structure size of the transformer oil tank reinforcing rib, enhances the anti-explosion ability of the oil tank mechanical structure, significantly reduces the risk of secondary combustion and explosion accidents caused by internal faults of the transformer, and has great practical application value in the industrial field. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 It is a flow chart of the transformer reinforcing rib size optimization method in the present invention;

[0040] Figure 2 It is a structural diagram of an oil-immersed transformer oil tank equipped with a reinforcing rib in the present invention;

[0041] Figure 3 For Figure 2 Schematic diagram at position A in

[0042] Figure 4 It is a finite element mesh division diagram in the present invention;

[0043] Figure 5The deformed cloud map of the fuel tank wall before the optimization of the stiffeners in the present invention;

[0044] Figure 6 The deformed cloud map of the fuel tank wall after the optimization of the stiffeners in the present invention;

[0045] Figure 7 The schematic diagram of the transformer stiffener size optimization system in the present invention;

[0046] In the figure, 1 is the oil conservator, 2 is the connecting pipe, 3 is the pressure relief valve, 4 is the stiffener, and 5 is the fuel tank. Specific implementation mode

[0047] The present invention will be described in detail below with reference to the accompanying drawings.

[0048] For the convenience of understanding the present invention, the present invention will be described more comprehensively and in detail in combination with embodiments. It should be noted that the present invention can be implemented in many different forms, and the embodiments described herein are only used to better explain the present invention and make the content of the present invention easier to understand thoroughly, and are not the only limitation.

[0049] A method for optimizing the size of the transformer stiffeners in the present invention realizes the optimization of the size of the stiffeners 4 by combining the accurate numerical calculation of the ANSYS (Analysis of Systems) finite element simulation platform and the surrogate function optimization, so as to maximize the explosion-proof performance of the transformer tank wall. The optimization method specifically includes the following steps:

[0050] (1) A pressure relief valve 3 and a connecting pipe 2 are provided on the oil-immersed transformer fuel tank 5. The top of the connecting pipe 2 is connected to the oil conservator 1, and the oil-immersed transformer fuel tank 5 is equipped with stiffeners 4.

[0051] Select the parameters to be optimized and the design space of the transformer stiffeners 4. In the design space, sampling samples are obtained. Among them, the parameters to be optimized of the transformer stiffeners 4 include the length of the stiffeners l , the height of the stiffeners h and the wall thickness of the stiffeners d ;

[0052] Specifically, according to the structural constraints of the transformer fuel tank 5, the design space of the parameters to be optimized of the stiffeners 4 is determined;

[0053] In the design space, the Monte Carlo sampling method is used for sample collection. The required number of random sample points are generated by using a normal distribution random number generator through formulas (1) and (2) to obtain sampling samples;

[0054] (1)

[0055] (2)

[0056] Among them, I is the theoretical value of integration, is the approximate value of integration, is the integrand of the input data, is the integrand of the i c th data sample in the sampling sample vector, x 0 is the input data, is the sampling interval, a is the left endpoint of the interval, b is the right endpoint of the interval, N c is the number of sampling samples, is the i c th data sample in the sampling sample vector, i c is the sample number, i c = 1, 2…, is the probability density function.

[0057] According to the number of sampling samples and the distribution of random sample points, the accuracy of generating the required number of random sample points is evaluated using the standard error.

[0058] (2)Based on the sampling samples, through the fluid-structure interaction model, the deformation simulation calculation of the transformer tank wall structure is carried out to obtain the deformation amount of the tank wall of the transformer oil tank 5; the fluid-structure interaction model is established through the following process:

[0059] Based on a sampling sample, a three-dimensional geometric model of the transformer oil tank 5 is built, including a fluid domain model and a solid domain model, and the mesh division of the three-dimensional geometric model of the transformer oil tank 5 is carried out;

[0060] At each moment, based on the bubble dynamics equation under oil arc fault, the relationship between the behavior of the arc-induced bubbles inside the transformer and the increase in oil pressure is described, providing the initial oil arc fault energy for the fluid-structure interaction model. Among them, the bubble dynamics equation under oil arc fault is as follows:

[0061] (3)

[0062] Among them, is the energy injected by the high-energy arc, is the dissipated energy, is the arc energy conversion coefficient, is the bubble radius, is the radial velocity of the bubble surface, is the radial acceleration of the bubble surface, is the specific heat ratio, is the bubble volume, is the internal energy of the bubble, is the radial distance, is the fluid domain pressure, is the fluid density.

[0063] Set the fluid domain model solver as the transient solver. Based on the initial oil arc fault energy at the current moment, perform fluid solution through the Reynolds-Averaged Navier-Stokes Equations (RANS) to obtain the fluid domain pressure distribution results and displacement data at the current moment. The Reynolds-Averaged Navier-Stokes Equations (RANS) are as follows:

[0064] (4)

[0065] Among them, is the fluid density, is the time, is the spatial coordinate x coordinate component, is the spatial coordinate y coordinate component, is the spatial coordinate z coordinate component, is the fluid viscosity, is the Reynolds-averaged velocity x coordinate component, is the Reynolds-averaged velocity y coordinate component, is the Reynolds-averaged velocity z coordinate component, is the Reynolds stress tensor, is the Reynolds-averaged pressure, is the partial differential symbol.

[0066] Set the material properties of the solid structure of the transformer tank 5. Set the contact surface between the fluid domain and the solid domain on the ANSYS finite element simulation platform. Set the solid domain model solver as the transient solver. Use the fluid domain pressure distribution results and displacement data at the current moment as the load of the solid domain model solver to solve the deformation of the wall of the transformer tank 5 at the current moment;

[0067] In the next moment, use the deformation of the wall of the transformer tank 5 at the previous moment as the input of the fluid domain model solver. Update the fluid domain boundary of the fluid domain model through the dynamic mesh method to complete the transfer of the pressure distribution results and displacement data between the fluid domain and the solid domain, realize the fluid-structure interaction solution of the deformation of the transformer tank wall structure, and record and save the solution results of the deformation of the wall of the transformer tank at the final moment.

[0068] Modify the dimensional parameters of the stiffener 4 according to the sampling samples, and calculate the next sampling sample until the calculations for all sampling samples are completed, obtaining the deformation amounts of the transformer tank wall for all sampling samples.

[0069] (3) Fit the non - linear function relationship between the parameters to be optimized of the stiffener 4 and the deformation amount of the transformer tank wall to obtain a surrogate function. The specific steps are as follows:

[0070] Record each sampling sample and the corresponding deformation amount of the transformer tank wall to form a surrogate function data set , , where N d is the total number of samples in the surrogate function data set, is the - th data sample in the surrogate function input data set, is the deformation amount of the transformer tank wall corresponding to the - th data sample in the surrogate function data set.

[0071] According to the surrogate function data set, select the Least Squares Support Vector Regression (LSSVR) model to construct the surrogate function. Among them, the objective function of the LSSVR problem:

[0072] (5)

[0073] (6)

[0074] Where is the objective function of the LSSVR problem, is the weight vector of the complexity of the least squares support vector regression model, is the function bias term, is the error vector for calculating the error, w T is the transposed vector, is the function regularization parameter, is the non - linear feature mapping relationship from the input data to the mathematical high - dimensional space, N d is the total number of samples in the surrogate function data set, is the - th data sample in the surrogate function input data set, is the deformation amount of the transformer tank wall corresponding to the - th data sample in the surrogate function data set.

[0075] Based on the objective function of the LSSVR (Least Squares Support Vector Machine For Regression) problem, the Lagrange function is introduced, and considering the dual optimization, a system of linear equations is obtained:

[0076] (7)

[0077] where, is the Lagrange multiplier, is an matrix with all elements being 1, is an matrix with all elements being 1, is the identity diagonal matrix, is the reciprocal of is the function regularization parameter, is the radial basis function matrix, and the elements of the radial basis function matrix are selected as the Sigmoid kernel function:

[0078] (8)

[0079] (9)

[0080] where, is the element of the radial basis function matrix, is the adjustable parameter of the function, is the bias parameter of the function, is the radial basis function, is the characteristic function corresponding to the Sigmoid kernel function, is the hyperbolic tangent function. is the th data sample in the input dataset of the surrogate function, is the th data sample in the input dataset of the surrogate function. , , N d is the total number of samples in the surrogate function dataset.

[0081] By optimizing the objective function of the LSSVR problem, that is, the solution of the LSSVR problem can be expressed as the surrogate function :

[0082] (10)

[0083] where, Input the data set for the surrogate function, is the weight vector of the complexity of the least squares support vector regression model, , w T is the transposed vector, is the eigenfunction corresponding to the Sigmoid kernel function, is the function bias term, n is the total number of input data samples, is the Lagrange multiplier, is the radial basis function, is the -th data sample in the input data set of the surrogate function, is the -th data sample in the input data set of the surrogate function, is the first data sample serial number, is the second data sample serial number, is different from .

[0084] (4) Based on the surrogate function, use the particle swarm optimization algorithm to perform parameter optimization to obtain the numerical values of the parameters to be optimized. According to the numerical values of the parameters to be optimized, determine the dimensions of the reinforcing rib 4. The goal of parameter optimization is to minimize the deformation of the transformer tank cavity wall. The specific steps are as follows:

[0085] a) Initialize the particle distribution and update the velocity;

[0086] b) By comparing the fitness function values of each particle, re-update the individual historical best position and the population global historical optimal position, and update the particle velocity according to the following formula and the particle position :

[0087] (11)

[0088] (12)

[0089] In the formula, is the inertia factor, is the iteration generation number, is the particle velocity of the particle at the ip -th generation, is the particle position of the particle at the ip -th generation, is the personal best position of the particle at the ip -th generation, is the global best position of all particles, c 1 is the learning factor from the individual best, c 2 is the learning factor from the global best; r1 is the first random number, r 2 is the second random number, and the first random number r 1 and the second random number r 2 are between 0 and 1, ip which is the particle sequence number.

[0090] c) If the number of iterations of the particle swarm optimization reaches the preset maximum number of iterations, or the change in the optimal value of the objective function is less than the preset value, it is determined that the particle swarm optimization algorithm meets the convergence condition, and the parameters to be optimized are output. Otherwise, step b) is continued.

[0091] Taking the oil-immersed transformer as the research object below, the dimension optimization of the reinforcing rib 4 of the transformer is carried out.

[0092] See Figure 1 , a method for optimizing the dimensions of the reinforcing rib of a transformer, comprising the following steps:

[0093] Step 1: As Figure 2 and Figure 3 shown, a pressure relief valve 3 and a connecting pipe 2 are provided on the oil tank 5 of the oil-immersed transformer. The top of the connecting pipe 2 is connected to the conservator 1. The oil tank 5 of the oil-immersed transformer is equipped with a reinforcing rib 4. The cross-section of the reinforcing rib 4 is a hollow rectangle. The length of the reinforcing rib l , the height of the reinforcing rib h and the wall thickness of the reinforcing rib d are the parameters to be optimized.

[0094] Step 2: Determine the design space of the parameters to be optimized for the reinforcing rib 4 according to the structural constraints of the oil tank 5 of the oil-immersed transformer.

[0095] Step 3: Use the Monte Carlo sampling method to collect samples in the design space, and use the standard error to evaluate the accuracy of generating the required number of random sample points, obtaining 100 sets of training set data and 30 sets of validation set data.

[0096] Step 4: Rely on the ANSYS finite element simulation platform to build a three-dimensional geometric model of the oil tank 5 of the transformer. The three-dimensional geometric model of the oil tank 5 of the transformer includes a solid domain model and a fluid domain model.

[0097] Step 5: Use the meshing software (Meshing software) to complete the finite element meshing of the solid domain model and the fluid domain model. The meshing result is as Figure 4 shown.

[0098] Step 6: Set the transient solver of the fluid domain model in the Fluent software (computational fluid dynamics software); the solution time is set to 80 ms, and the solution step size is 0.1 ms.

[0099] Step 7: Describe the bubble behavior using user-defined functions (UDF) and the bubble dynamics equation, and update the fluid domain boundary using the dynamic mesh method.

[0100] Step 8: Set the solid domain model in Transient Structural software. The material of the tank wall is Q235 steel, with a density of 7850 kg / m 3 , a Young's modulus of 203 GPa, a yield stress of 235 MPa, and a tensile strength of 460 MP.

[0101] Step 9: Set the contact surface between the fluid domain and the solid domain on the ANSYS finite element simulation platform to achieve the transfer of force and displacement data between the fluid domain and the solid field; use the transient solver to solve the fluid-structure interaction model under internal arc faults and solve for the deformation of the tank wall of the transformer tank 5.

[0102] Step 10: Traverse the training set data, record and save the deformation of the tank wall of the transformer tank 5 until all the sampling samples in the training set are calculated, and form a surrogate function dataset. .

[0103] Step 11: Select the least squares support vector regression model (LSSVR) to construct a surrogate function based on the surrogate function dataset.

[0104] By optimizing the objective function of the LSSVR problem, the solution of the LSSVR problem can be expressed as a surrogate function :

[0105] (10)

[0106] where is the input dataset of the surrogate function, is the weight vector of the complexity of the least squares support vector regression model, , w T is the transposed vector, is the characteristic function corresponding to the Sigmoid kernel function, is the function bias term, n is the total number of input data samples, is the Lagrange multiplier, is the radial basis function, is the th data sample in the input dataset of the surrogate function, is the th data sample in the input dataset of the surrogate function, is the serial number of the first data sample, is the serial number of the second data sample, which is different from

[0107] Step 12: Verify the surrogate function with the validation set data, correct the surrogate function parameters and increase the sampling rate to meet the error condition, that is, the calculated error is less than the error upper limit.

[0108] Step 13: The objective function of the LSSVR problem is to minimize the maximum deformation of the wall of the transformer tank 5 , and the mathematical forms of the objective function and inequality constraints of the LSSVR problem are:

[0109] (13)

[0110] (14)

[0111] where is the weight vector of the optimization algorithm.

[0112] Step 14: Select the particle swarm optimization algorithm to optimize the surrogate function. The specific steps are as follows:

[0113] Initialize the particles and velocities: update the individual historical best position and the population global historical optimal position by comparing the fitness function values of each particle,

[0114] Update the particle velocity according to the following formula and the particle position :

[0115] (11)

[0116] (12)

[0117] In the formula, is the inertia factor, is the iteration generation, is the particle velocity of the ip -th generation of the particle, is the particle position of the ip -th generation of the particle, is the personal best position of the ip -th generation of the particle, is the global best position of all particles, c 1 is the learning factor from the personal best, c 2 is the learning factor from the global best; r 1 is the first random number, r 2 is the second random number. The first random number r 1 and the second random number r2 is between 0 and 1, ip which is the particle sequence number.

[0118] If the maximum number of iterations has not reached the maximum number of iterations, and the change in the optimal value of the objective function is greater than the set threshold, then continue to execute the previous step. If the convergence condition is met, output the optimization scheme of the structural parameters of the transformer tank wall stiffener 4.

[0119] The specific output of the surrogate function for the optimization result of the structural dimensions of the stiffener 4 is: the optimal length of the stiffener l , the height of the stiffener h and the wall thickness of the stiffener d are 49.79 mm, 34.81 mm and 14.92 mm respectively. The total energy released by the arc fault caused by the internal fault of the transformer before and after optimization is the same, which is 10 MJ.

[0120] Figure 5 and Figure 6 show the comparison of the deformation distribution of the transformer tank wall before and after the optimization of the structural dimensions of the stiffener 4. Before optimizing the structural dimensions of the stiffener 4, the displacement deformation of the wall of the oil tank 5 is concentrated in the middle of the wall. This difference indicates that the oil pressure distribution on different structural regions of the oil-immersed transformer is uneven under the oil-gas arc fault. After optimizing the structural dimensions of the stiffener 4, the displacement deformation distribution of the tank wall is more uniform. The maximum deformation displacement of the wall of the oil tank 5 is reduced from 15.1 cm to 11.5 cm, with a decrease of 23.8%. The results show that this optimization method enhances the explosion-proof performance of the oil-immersed transformer oil tank 5 and significantly reduces the risk of transformer explosion accidents.

[0121] Refer to Figure 7 , in an embodiment of the present invention, a transformer stiffener size optimization system is provided, including:

[0122] A module for determining parameters to be optimized and design space, which is used to determine the parameters to be optimized and the design space of the stiffener 4 of the transformer; among them, the parameters to be optimized of the stiffener 4 include the length of the stiffener, the height of the stiffener and the wall thickness of the stiffener;

[0123] A sample collection module, which is used to collect samples in the design space to obtain sampling samples;

[0124] A simulation calculation module, which is used to perform simulation calculations on the deformation of the transformer tank wall structure based on the sampling samples through a fluid-structure interaction model to obtain the deformation amount of the wall of the transformer oil tank 5;

[0125] A fitting module, which is used to fit the non-linear function relationship between the parameters to be optimized of the stiffener 4 of the transformer and the deformation amount of the wall of the transformer oil tank 5 to obtain a surrogate function;

[0126] A parameter optimization module, which is used to perform parameter optimization based on the surrogate function to obtain the numerical values of the parameters to be optimized and determine the dimensions of the stiffener 4 of the transformer.

[0127] In one embodiment of the present invention, 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 processor executes the computer program, the method for optimizing the size of the transformer reinforcing ribs is implemented.

[0128] In one embodiment of the present invention, a computer-readable storage medium is provided. The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method for optimizing the size of the transformer reinforcing ribs is implemented.

[0129] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a system, or a computer program product. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, optical storage, etc.) containing computer-usable program code.

[0130] The present invention is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to the embodiments of the present invention. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, as well as the combination of flows and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, such that the instructions executed by the processor of the computer or other programmable data processing devices generate means for implementing the specified functions in Figure 1 one or more of the flows Figure 1 or blocks or the combination of blocks.

[0131] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, such that the instructions stored in the computer-readable memory generate a manufactured article including instruction means for implementing the specified functions in Figure 1 one or more of the flows Figure 1 or blocks or the combination of blocks.

[0132] These computer program instructions can also be loaded onto a computer or other programmable data processing device, such that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process. Thus, the instructions executed on the computer or other programmable device provide means for implementing the specified functions in Figure 1Steps of a process or multiple processes and / or boxes Figure 1 Steps of the functions specified in a box or multiple boxes.

[0133] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them. Although the present invention has been described in detail with reference to the above embodiments, those of ordinary skill in the art should understand that: modifications or equivalent replacements can still be made to the specific embodiments of the present invention, and any modifications or equivalent replacements 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 method for optimizing the size of the transformer reinforcing rib, characterized in that, It includes the following steps: Determine the parameters to be optimized and the design space of the transformer stiffener (4); among them, the parameters to be optimized of the transformer stiffener (4) include the length of the stiffener (4), the height of the stiffener (4), and the wall thickness of the stiffener (4); Collect samples in the design space to obtain sampling samples; Based on the sampling samples, through the fluid-structure interaction model, perform simulation calculations on the deformation of the transformer tank wall structure to obtain the deformation amount of the transformer oil tank wall; Fit the non-linear function relationship between the parameters to be optimized of the transformer stiffener (4) and the deformation amount of the transformer oil tank wall to obtain a surrogate function; Based on the surrogate function, perform parameter optimization to obtain the numerical values of the parameters to be optimized, and determine the size of the transformer stiffener (4); Among them, through the fluid-structure interaction model, perform simulation calculations on the deformation of the transformer tank wall structure to obtain the deformation amount of the transformer oil tank wall, including the following steps: Based on a sampling sample, build a three-dimensional geometric model of the transformer oil tank, including a fluid domain model and a solid domain model; Solve the fluid domain model after updating the fluid domain boundary to obtain the fluid domain pressure distribution results and displacement data at the current moment; Set the material properties of the solid structure of the transformer oil tank and the contact surface between the fluid domain and the solid domain; Based on the fluid domain pressure distribution results and displacement data at the current moment, solve the solid domain model to obtain the deformation amount of the transformer oil tank wall at the current moment; Use the dynamic mesh method to update the fluid domain boundary in the fluid domain model at the next moment, complete the transfer of the pressure distribution results and displacement data between the fluid domain and the solid domain, and obtain the deformation amount of the transformer oil tank wall at the final moment; Perform calculations for the next sampling sample until the calculations for all sampling samples are completed, and obtain the deformation amounts of the transformer oil tank walls for all sampling samples.

2. The method for optimizing the size of the transformer reinforcing rib according to claim 1, characterized in that The sampling in the design space to obtain sampling samples is performed using the Monte Carlo sampling method.

3. The method for optimizing the size of the transformer reinforcing rib according to claim 1, characterized in that The fitting of the non-linear function relationship between the parameters to be optimized of the transformer stiffener (4) and the deformation amount of the transformer oil tank wall to obtain a surrogate function includes the following steps: Form a surrogate function data set with each sampling sample and the corresponding deformation amount of the transformer oil tank wall; Based on the surrogate function data set, construct a surrogate function through the least squares support vector regression model.

4. The method for optimizing the size of the transformer reinforcing rib according to claim 1, wherein The surrogate function is: Among them, is the proxy function, is the input data set of the proxy function, is the weight vector of the complexity of the least squares support vector regression model, w T is the transposed vector, is the characteristic function corresponding to the Sigmoid kernel function, is the function bias term, n is the total number of input data samples, is the Lagrange multiplier, is the radial basis function, is the -th data sample in the input data set of the proxy function, is the -th data sample in the input data set of the proxy function, is the first data sample number, is the second data sample number, and are different.

5. The method for optimizing the size of the transformer reinforcing rib according to claim 1, characterized in that The parameter optimization based on the surrogate function to obtain the numerical values of the parameters to be optimized is performed using the particle swarm optimization algorithm.

6. The method for optimizing the size of the transformer stiffener according to claim 1, characterized in that The goal of parameter optimization based on the surrogate function is to minimize the deformation of the transformer oil tank cavity wall.

7. A transformer rib size optimization system, characterized in that, It includes: A module for determining parameters to be optimized and the design space, which is used to determine the parameters to be optimized and the design space of the transformer stiffener (4); among them, the parameters to be optimized of the transformer stiffener (4) include the length of the stiffener (4), the height of the stiffener (4), and the wall thickness of the stiffener (4); A sample collection module, which is used to collect samples in the design space to obtain sampling samples; A simulation calculation module, which is used to perform simulation calculations on the deformation of the transformer tank wall structure based on the sampling samples through the fluid-structure interaction model to obtain the deformation amount of the transformer oil tank wall; A fitting module, configured to fit the non-linear function relationship between the parameters to be optimized of the transformer stiffener (4) and the deformation amount of the transformer tank wall to obtain a surrogate function; A parameter optimization module, configured to perform parameter optimization based on the surrogate function to obtain the numerical values of the parameters to be optimized and determine the dimensions of the transformer stiffener (4); Wherein, through a fluid-structure interaction model, a simulation calculation of the structural deformation of the transformer tank wall is performed to obtain the deformation amount of the transformer tank wall, including the following steps: Based on a sampling sample, a three-dimensional geometric model of the transformer tank is built, including a fluid domain model and a solid domain model; Solve the fluid domain model after updating the fluid domain boundary to obtain the fluid domain pressure distribution result and displacement data at the current moment; Set the material properties of the solid structure of the transformer tank and the contact surface between the fluid domain and the solid domain; Based on the fluid domain pressure distribution result and displacement data at the current moment, solve the solid domain model to obtain the deformation amount of the transformer tank wall at the current moment; Use the dynamic mesh method to update the fluid domain boundary in the fluid domain model at the next moment, complete the transfer of the pressure distribution result and displacement data between the fluid domain and the solid domain, and obtain the deformation amount of the transformer tank wall at the final moment; Perform the calculation of the next sampling sample until the calculation of all sampling samples is completed, and obtain the deformation amounts of the transformer tank walls of all sampling samples.

8. An electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein, When the processor executes the computer program, it implements the transformer stiffener size optimization method according to any one of claims 1 to 6.

9. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the transformer stiffener size optimization method according to any one of claims 1 to 6.

Images