A linear phase-shifting transformer optimization design method, system, device and terminal

Abstract

The present invention belongs to the technical field of transformers and discloses a linear phase-shifting transformer optimization design method, system, device and terminal. Based on the equivalent circuit of the linear phase-shifting transformer, several key design parameters are extracted as optimization variables; the optimization design constraints and objective functions applicable to the linear phase-shifting transformer model are defined to establish an optimization design model; a differential evolution algorithm is used for parameter optimization design, and the linear phase-shifting transformer is optimized using the equivalent circuit. The present invention proposes a method for optimizing the design of a linear phase-shifting transformer based on an equivalent circuit model and utilizing the advantages of the differential evolution algorithm in information interaction and intelligent learning. Under the condition of meeting design requirements and conditional constraints, the present invention uses the algorithm's optimization capability to obtain the efficiency, voltage regulation rate, weight and other characteristics of the linear phase-shifting transformer, which can reach the single value optimum or comprehensive optimum within the feasible domain, thereby reducing the core loss of the linear phase-shifting transformer.

Description

Technical Field

[0001] The present invention belongs to the technical field of transformers, and in particular relates to a linear phase-shifting transformer optimization design method, system, equipment and terminal. Background Art

[0002] As a new type of phase-shifting transformer, the linear phase-shifting transformer offers advantages over traditional core-type phase-shifting transformers, such as the ability to shift phase at any angle, good scalability, and ease of modularization. It also provides electrical isolation, isolating the power supply from the load, making it suitable for high-power multiple inverter and multiple rectifier systems on ships. When used in a multiple inverter system, the linear phase-shifting transformer's primary function is to synthesize the square wave voltages output by multiple inverters, producing a highly sinusoidal three-phase, multi-step voltage waveform. This eliminates low-order harmonics and enables a high-quality output voltage waveform even at low switching frequencies.

[0003] Traditional linear phase-shifting transformers suffer from performance issues such as high losses (core loss and winding copper loss), low overall efficiency, high voltage regulation, and heavy core weight. Traditional optimization methods for linear phase-shifting transformers tend to converge to local minima, making it difficult to achieve a global optimal solution and dealing with fuzzy factors and discrete variables.

[0004] Linear phase-shifting transformers are complex static induction devices consisting of an iron core and windings. Optimizing performance parameters requires considering numerous nonlinear and discrete relationships. Furthermore, the optimal design of linear phase-shifting transformers is subject to fuzzy factors and the difficulty of accurately expressing these relationships. From the perspective of the entire optimization process, the optimal design of linear phase-shifting transformers can be mathematically formulated as a constrained, multivariable, and nonlinear mixed programming problem, with the objective function exhibiting multiple extreme points.

[0005] The equivalent circuit model of the linear phase-shifting transformer can be used to describe steady-state electromagnetic relationships, analyze steady-state operation characteristics, and provide guidance for the design optimization of the linear phase-shifting transformer. Through the equivalent circuit model, the parameters that need to be optimized for the linear phase-shifting transformer can be selected, the optimization design constraints and objective functions applicable to the linear phase-shifting transformer model can be defined, and the optimization design model can be established.

[0006] The performance of an algorithm determines whether the global optimal solution can be effectively found in an optimization design. Therefore, selecting the appropriate optimization algorithm is crucial for the multi-objective optimization design of linear phase-shifting transformers. Currently, these algorithms can be primarily categorized into two types: gradient-based optimization algorithms and intelligent optimization algorithms. Gradient-based optimization algorithms can be further divided into conjugate gradient optimization algorithms and sequential quadratic programming algorithms. There are many different types of intelligent optimization algorithms, with the most commonly used being evolutionary algorithms, particle swarm optimization algorithms, non-dominated sorting genetic algorithms, and multi-objective genetic algorithms.

[0007] The differential evolution algorithm is an intelligent optimization algorithm derived from natural laws such as survival of the fittest and genetic variation. The algorithm defines a feasible region for each variable in a continuous space and searches and learns within that region to optimize the objective function. The differential evolution algorithm includes three strategies: mutation, crossover, and selection. A variety of complex function test sets and real-world engineering applications with constraints have demonstrated the effectiveness and feasibility of the differential evolution algorithm and its variants.

[0008] Through the above analysis, the problems and defects of the existing technology are as follows:

[0009] (1) Traditional linear phase-shifting transformers have performance problems such as large losses (core loss and winding copper loss), low overall efficiency, large voltage regulation rate, and heavy core weight.

[0010] (2) The traditional optimization methods applied to linear phase-shifting transformers have the disadvantages of being easily converged to local optimal points, being difficult to guarantee a global optimal solution, and being difficult to handle fuzzy factors and discrete variables.

[0011] (3) In the optimization problem of performance parameter selection of linear phase-shifting transformers, many nonlinear and discrete relationships need to be considered, and there are fuzzy factors and difficulties in accurately quantitatively expressing relationships in the optimization design. Summary of the Invention

[0012] In response to the problems existing in the existing linear phase-shifting transformer design technology, the present invention provides a linear phase-shifting transformer optimization design method, system, device and terminal, and in particular relates to a linear phase-shifting transformer optimization design method, system, device and terminal based on an equivalent circuit model.

[0013] The present invention is implemented as follows: a linear phase-shifting transformer optimization design method, the linear phase-shifting transformer optimization design method includes: based on the equivalent circuit of the linear phase-shifting transformer, extracting several key design parameters as optimization variables; defining optimization design constraints and objective functions applicable to the linear phase-shifting transformer model, and establishing an optimization design model; using a differential evolution algorithm to perform parameter optimization design, and using the equivalent circuit to optimize the linear phase-shifting transformer.

[0014] Furthermore, the linear phase-shifting transformer optimization design method also includes: inputting algorithm parameters and initializing the parameter matrix; selecting the global optimal solution according to the objective function; judging whether Fes is less than Max_Fes, and generating a matrix of test vectors; performing boundary processing and generating a matrix of cross vectors; selecting a better vector from the test vectors and the cross vectors, and finally selecting the global optimal solution according to the objective function value.

[0015] Furthermore, the linear phase-shifting transformer optimization design method includes the following steps:

[0016] Step 1: Determine the optimized design model: Analyze the equivalent circuit model of the linear phase-shifting transformer and extract the parameters to be optimized in the equivalent circuit model;

[0017] Step 2: Use the differential evolution algorithm for optimization design: Use the differential evolution algorithm to optimize the various parameters in the equivalent circuit model and finally obtain the optimal output result.

[0018] Furthermore, the determination of the optimized design model in step 1 includes:

[0019] (1) Optimization parameter selection

[0020] In the optimization design of a linear phase-shifting transformer, a total of five design parameters and / or variables are selected for optimization. The parameter optimization range is within the effective feasible domain, where the feasible domain is the value range. The selected design objects are arranged into groups in a fixed order, and each group of sequentially arranged design objects constitutes a vector. Based on the original design data of the linear phase-shifting transformer and combined with the experimental prototype parameters, the value range of the parameter objects is optimized.

[0021] (2) Optimization design constraints

[0022] Conditional constraints refer to the actual conditions that the linear phase-shifting transformer designed with parameters within the feasible region meets during operation. Based on references and the actual design of linear phase-shifting transformers, the following constraints are defined:

[0023] B tooth ≤1.4T;

[0024] a / τ≥0.7;

[0025] R sf ≤0.7;

[0026] gd≥0.2mm;

[0027] Among them, B tooth Refers to the tooth flux density amplitude, which is specified to be no greater than 1.4T; a / τ is the core thickness to pole pitch ratio, which is used to constrain the pole pitch, prevent it from being too wide, and also reduce the length of the winding end; R sf represents the slot fill rate, which is set to 0.7; g is the air gap length, d is the secondary core thickness, and the minimum value of the air gap between the primary and secondary core surfaces is set to 0.2 mm.

[0028] (3) Determination of the objective function

[0029] The slot fill rate is taken as the optimization target of the linear phase-shifting transformer. The efficiency, power factor, linear phase-shifting transformer body weight and slot fill rate are selected as multiple optimization targets to construct a comprehensive multi-objective optimization function to guide the optimization design process of the linear phase-shifting transformer. The constructed comprehensive multi-objective function expression is:

[0030]

[0031] Among them, η represents efficiency, PF is power factor, R sf is the slot fill rate, Weight is the tooth weight; k i Represents the weight coefficient of the optimization target, i=1,2,3, and can only be 0 or 1.

[0032] Furthermore, the value range of the optimization parameter object in step (1) includes:

[0033] The winding methods include full-pitch stacked winding on the primary side and long-short pitch winding on the secondary side;

[0034] The minimum core stack thickness is 40mm and the maximum is 100mm;

[0035] The minimum value of the slot width to slot pitch ratio is 0.3 and the maximum value is 0.7;

[0036] The minimum tooth height is 10mm and the maximum is 50mm;

[0037] The minimum current density is 3×10 6 A / mm 2 , the maximum value is 6×10 6 A / mm 2 ;

[0038] The minimum value of the air gap length is 0.02mm and the maximum value is 0.05mm.

[0039] Furthermore, the optimization design using the differential evolution algorithm in step 2 includes:

[0040] (1) Initialization settings

[0041] In the initialization phase, the differential evolution algorithm randomly generates N V D-dimensional vectors, where NV represents the number of vectors in the matrix, which is set to 50. In one iteration, information interaction and learning are performed through 50 different solution vectors; D represents the number of variables contained in the vector, which refers to the number of design parameters contained in the vector, and is set to 5. In the initial solution matrix, the j-th dimension variable in the i-th vector is generated using the following formula:

[0042]

[0043] Among them, i represents the i-th vector in the matrix, i∈[1,N V ]; j represents the j-th dimension variable in the vector, j∈[1,D]; represents the j-th dimension variable of the i-th vector in the initial solution matrix, represents the j-th optimization variable of the i-th group of parameters; L(j) represents the lower boundary of the j-th dimension of the vector; U(j) represents the upper boundary of the j-th dimension of the vector; rand represents a random number uniformly distributed between 0 and 1; the generated initial solution matrix is:

[0044]

[0045] The initial solution matrix contains N V vectors, each with D variables; each vector represents a candidate solution, and the D variables in the candidate solution represent D design parameter values. In each iteration, information exchange and intelligent learning are achieved between different vectors, allowing each vector to evolve in a better direction. Ultimately, the optimal vector is output, whose D variables represent the optimal values ​​of the linear phase-shifting transformer design parameters.

[0046] (2) Mutation strategy

[0047] Mutation strategy in N V Perform differential operations between different vectors X to generate mutation vectors V. The order and position of the parameters contained in each vector remain unchanged during the optimization process, and only the values ​​change continuously through learning. For the differential evolution algorithm, the mutation strategy calculation formula "DE / rand / 1" is:

[0048]

[0049] Among them, V i t Represents the i-th mutation vector generated in the t-th iteration, i∈[1,N V ]; r1, r2, r3∈[1, N V ] are three different integers; F i t is the i-th scaling factor in the t-th iteration, which is used to control the weight of the differential operation and generates the i-th mutation vector, F i t The value is set between 0 and 2.

[0050] The new mutation vector matrix generated by the mutation strategy is:

[0051]

[0052] (3) Boundary processing

[0053] Pull the dimension beyond the boundary back to the boundary. The mathematical expression is:

[0054]

[0055] Where t represents the tth iteration; i represents the i-th vector in the matrix, i∈[1,N V ]; j represents the j-th dimension variable in a vector, j∈[1,D].

[0056] (4) Crossover strategy

[0057] After boundary processing, the algorithm crossover strategy is executed. The crossover strategy is to perform random replication of a certain dimension between different vectors X and crossover vectors V within certain rules to generate a new trial vector U.

[0058] The new trial vector stores information about both the old vector X and the new crossover vector V, and combines this information in a certain way to explore new potential ranges. The crossover strategy allows the differential evolution algorithm to further enhance its global search capabilities within the feasible domain. The crossover strategy formula is:

[0059]

[0060] in, represents the j-th dimension variable of the i-th trial vector in the t-th iteration; CR∈[0,1] represents the crossover probability, which is a constant to be defined; rand represents a random number uniformly distributed between 0 and 1; j rand Represents a random integer randomly selected from the range [1,2,…,D].

[0061] The trial vector matrix is ​​expressed as:

[0062]

[0063] For the optimization design of linear phase-shifting transformer, the crossover strategy is used to recombine different optimization design parameters to analyze the impact of the new design parameter combination on the performance of the linear phase-shifting transformer.

[0064] (5) Select strategy

[0065] The differential evolution algorithm uses the optimal criterion to select the next generation parameter vector X. The test vector U and the parameter vector X of the previous generation are evaluated, and the vector with better fitness is selected as the parameter vector of the next generation.

[0066] The selection strategy is performed as follows:

[0067]

[0068] Where f[] represents the fitness value of the vector. In the linear phase-shifting transformer optimization design problem, the fitness value is the objective function value calculated by substituting the parameters into the equivalent circuit. Since the linear phase-shifting transformer optimization problem is to maximize the objective function, the optimal criterion selects the vector with the larger objective function value among the test vectors or parameter vectors as the next generation parameter vector.

[0069] Another object of the present invention is to provide a linear phase-shifting transformer optimization design system using the linear phase-shifting transformer optimization design method. The linear phase-shifting transformer optimization design system includes:

[0070] An optimization design model building module is used to determine the optimization design model by optimizing parameter selection, optimizing design constraints, and determining the objective function;

[0071] The transformer optimization design module is used to optimize the design of linear phase-shifting transformers using the differential evolution algorithm, including initialization settings, mutation strategy, boundary processing, crossover strategy and selection strategy.

[0072] Another object of the present invention is to provide a computer device, which includes a memory and a processor. The memory stores a computer program. When the computer program is executed by the processor, the processor executes the steps of the linear phase-shifting transformer optimization design method.

[0073] Another object of the present invention is to provide a computer-readable storage medium storing a computer program, which, when executed by a processor, causes the processor to execute the steps of the linear phase-shifting transformer optimization design method.

[0074] Another object of the present invention is to provide an information data processing terminal, which is used to implement the linear phase-shifting transformer optimization design system.

[0075] In combination with the above technical solutions and the technical problems solved, the advantages and positive effects of the technical solutions to be protected by the present invention are as follows:

[0076] First, in view of the technical problems existing in the above-mentioned prior art and the difficulty of solving these problems, we closely combine the technical solutions to be protected by the present invention and the results and data during the research and development process, and conduct a detailed and in-depth analysis of how the technical solutions of the present invention solve the technical problems and some creative technical effects brought about by solving the problems. The specific description is as follows:

[0077] The linear phase-shifting transformer optimization design method provided by the present invention is based on the equivalent circuit of the linear phase-shifting transformer, extracts several key design parameters as optimization variables, defines optimization design constraints and objective functions applicable to the linear phase-shifting transformer model, and establishes an optimization design model; adopts a differential evolution algorithm with strong global search performance for parameter optimization design, and utilizes the accuracy of the equivalent circuit and the optimization capability of the efficient algorithm to accurately and efficiently optimize the design of the linear phase-shifting transformer.

[0078] This paper proposes an optimization design for linear phase-shifting transformers based on an equivalent circuit model, leveraging the information exchange and intelligent learning capabilities of a differential evolution algorithm. While meeting design requirements and constraints, the algorithm's optimization capabilities enable the linear phase-shifting transformer to achieve single-value or combined optimality within its feasible domain for characteristics such as efficiency, voltage regulation, and weight.

[0079] The design of the linear phase-shifting transformer according to the present invention can bring the following advantages:

[0080] 1. Reduce the core loss of the linear phase-shifting transformer and improve efficiency;

[0081] 2. The voltage regulation rate and output harmonic content are reduced, and the performance of the linear phase-shifting transformer is optimized;

[0082] 3. After optimization, the core weight of the linear phase-shifting transformer is reduced, and the production cost is reduced.

[0083] Second, considering the technical solution as a whole or from the perspective of the product, the technical effects and advantages of the technical solution to be protected by the present invention are described in detail as follows:

[0084] This invention optimizes the design of a linear phase-shifting transformer based on an equivalent circuit model and an optimized design model, employing a differential evolution algorithm. This optimization method reduces approximations and assumptions in the design, reduces reliance on design experience, conserves computing resources, and lowers production costs.

[0085] Third, as auxiliary evidence of the invention's creativity, it is also reflected in the following important aspects:

[0086] (1) The expected benefits and commercial value of the technical solution of the present invention after transformation are:

[0087] The linear phase-shifting transformer has a large core mass and uses more windings. According to the design objectives, the linear phase-shifting transformer is optimized using the differential evolution algorithm, which can effectively improve efficiency, reduce losses and lower production costs.

[0088] (2) The technical solution of the present invention fills the technical gap in the industry at home and abroad:

[0089] There have been no reported cases in China of using differential evolution algorithm to optimize the design of linear phase-shifting transformers. This method fills the gap in the optimization design of linear phase-shifting transformers in China and achieves a breakthrough from scratch.

[0090] (3) The technical solution of the present invention solves the technical problems that people have been eager to solve but have never been able to solve successfully:

[0091] Solve the problem of fuzzy factors being difficult to quantify and discrete variables being difficult to handle in the design process of linear phase-shifting transformers

[0092] (4) The technical solution of the present invention overcomes technical prejudice:

[0093] It overcomes the technical bias of traditional linear phase-shifting transformer optimization methods that are prone to converge to local optimal points and are difficult to ensure a global optimal solution. BRIEF DESCRIPTION OF THE DRAWINGS

[0094] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments of the present invention. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0095] Figure 1 This is a flow chart of the linear phase-shifting transformer optimization design method provided by an embodiment of the present invention;

[0096] Figure 2 This is a schematic diagram of a linear phase-shifting transformer optimization design method provided by an embodiment of the present invention;

[0097] Figure 3 This is a T-type equivalent circuit diagram of a linear phase-shifting transformer provided by an embodiment of the present invention;

[0098] Figure 4 is a schematic diagram of a mutation strategy provided by an embodiment of the present invention;

[0099] Figure 5 This is a schematic diagram of a crossover strategy provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0100] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0101] In view of the problems existing in the prior art, the present invention provides a linear phase-shifting transformer optimization design method, system, device and terminal. The present invention is described in detail below with reference to the accompanying drawings.

[0102] 1. Explanatory Examples In order to enable those skilled in the art to fully understand how to implement the present invention, this section provides an illustrative example that expands on the technical solution.

[0103] like Figure 1 As shown, the linear phase-shifting transformer optimization design method provided by the embodiment of the present invention includes the following steps:

[0104] S101, based on the equivalent circuit of linear phase-shift transformer, extracts several key design parameters as optimization variables;

[0105] S102, defining optimization design constraints and objective functions applicable to a linear phase-shifting transformer model, and establishing an optimization design model;

[0106] S103, using a differential evolution algorithm to perform parameter optimization design, and using an equivalent circuit to optimize the design of the linear phase-shifting transformer.

[0107] As a preferred embodiment, Figure 2 As shown, the linear phase-shifting transformer optimization design method provided by the embodiment of the present invention specifically includes the following steps:

[0108] First, the equivalent circuit model of the linear phase-shifting transformer is analyzed to extract the parameters that need to be optimized in the model; secondly, the differential evolution algorithm is used to optimize the parameters; finally, the optimal output result is obtained.

[0109] 1. Determination of the optimized design model

[0110] 1. Optimize parameter selection.

[0111] In the linear phase-shifting transformer optimization problem of this invention, five design parameters (variables) are selected for optimization. The parameter optimization range must be within the valid feasible domain, also known as the value range. In the linear phase-shifting transformer optimization design, the selected design objects are arranged into groups in a fixed order. Each group of sequentially arranged design objects constitutes a vector.

[0112] According to the original design data of the linear phase-shifting transformer and combined with the experimental prototype parameters, the specific value range of the optimized parameter object is shown in Table 1.

[0113] Table 1 Value range of optimization parameter objects

[0114]

[0115] 2. Optimize design constraints.

[0116] Conditional constraints refer to the actual conditions that must be met during the operation of a linear phase-shifting transformer designed within the feasible domain. When optimizing the design, key and non-negligible conditions are extracted from the complex electromagnetic relationships. The constraints of the optimized design are rationally and comprehensively considered to ensure that the optimized design results are close to the actual situation. Based on references and the actual design of linear phase-shifting transformers, the present invention defines the following constraints:

[0117] B tooth ≤1.4T

[0118] a / τ≥0.7

[0119] R sf ≤0.7

[0120] gd≥0.2mm

[0121] Among them, B tooth Refers to the tooth magnetic flux amplitude, a is half of the primary core thickness, τ is the pole pitch, R sf represents the slot fill rate, g is the air gap length, and d is the secondary core thickness.

[0122] The present invention stipulates that the tooth flux density amplitude cannot exceed 1.4T. The core stack thickness to pole pitch ratio constrains the pole pitch, preventing excessive pole width and reducing the length of the winding ends. In actual projects, the slot fill factor limit is approximately 0.75, typically set at 0.7 to facilitate manufacturing. Furthermore, due to the relatively small air gap in linear induction linear phase-shifting transformers, the present invention sets a minimum air gap of 0.2mm between the primary and secondary surfaces of the core.

[0123] 3. Determination of objective function.

[0124] Improving the efficiency and power factor of the linear phase-shifting transformer is the main optimization goal of the present invention. Secondly, reducing the weight of the body can reduce costs, improve power density, and facilitate installation and disassembly. Therefore, the tooth weight of the linear phase-shifting transformer is also one of the optimization indicators. In order to be close to the actual engineering design, the present invention takes the slot fill rate as the optimization target to make the slot utilization rate of the optimized design linear phase-shifting transformer higher. This is also a special requirement of the optimization method proposed by the present invention. In summary, the present invention selects multiple optimization targets of efficiency, power factor, linear phase-shifting transformer body weight and slot fill rate to construct a comprehensive multi-objective optimization function to guide the optimization design process of the linear phase-shifting transformer.

[0125] The comprehensive multi-objective function expression constructed by the present invention is:

[0126]

[0127] Among them, η represents efficiency, PF is power factor, and Weight is tooth weight. i (i=1,2,3) represents the weight coefficient of the optimization target, which can only be 0 or 1

[0128] 2. Using Differential Evolution Algorithm for Optimization Design

[0129] This paper uses an advanced differential evolution algorithm to optimize the design of linear phase-shifting transformers. Differential evolution algorithms often work on minimizing, nonlinear, and non-differentiable continuous spatial functions. The strategies and calculation rules of differential evolution algorithms include mutation, crossover, selection, and boundary processing.

[0130] 1. Perform initial settings.

[0131] This paper studies the linear phase-shifting transformer optimization problem. Five design objects are selected for optimization and arranged into groups in a fixed order to form a vector. For each group, any parameter combinations that cannot vary within the feasible region are gradually discarded during the algorithm iteration process, ensuring that all design results are realistic and consistent with actual conditions.

[0132] In the initialization phase, the differential evolution algorithm randomly generates N V D-dimensional vectors, where NV represents the number of vectors in a matrix. In this problem, it is set to 50, which means that information interaction and learning are carried out through 50 different solution vectors in one iteration. D represents the number of variables contained in a vector, which here refers to the number of design parameters contained in a vector. Since the present invention contains 5 optimized design parameters, D is set to 5 in the present invention. In an initial solution matrix, the j-th dimension variable in the i-th vector is generated using the following formula:

[0133]

[0134] Among them, i represents the i-th vector in the matrix, i∈[1,N V ], j represents the j-th dimension variable j∈[1,D]. Therefore, represents the jth variable of the i-th vector in the initial solution matrix, and represents the jth optimization variable of the i-th group of parameters. L(j) represents the lower bound of the j-th dimension of the vector, U(j) represents the upper bound of the j-th dimension of the vector, and rand represents a random number uniformly distributed between 0 and 1. Because the initial solution is randomly generated in the search space, the search area is not limited to a local area but is dispersed globally, making it more likely to explore the potential space in the feasible region.

[0135] Therefore, the initial solution matrix generated by this method is:

[0136]

[0137] The matrix contains N V vectors, each with D variables. Each vector represents a candidate solution, and the D variables in that candidate solution represent D design parameter values. In each iteration, information exchange and intelligent learning occur between the different vectors, allowing each vector to evolve towards a better solution. Ultimately, an optimal vector is output, whose D variables represent the optimal values ​​of the linear phase-shifting transformer design parameters.

[0138] 2. Mutation strategy.

[0139] Mutation strategy in N V Differential operations are performed between different vectors X to generate mutation vectors V. The order and position of the parameters contained in each vector remain unchanged during the optimization process; only their values ​​change continuously through learning. In the mutation strategy, the frequent and random information interaction between vectors creates the conditions for the algorithm to search for the optimal design parameters in the feasible domain. For differential evolution algorithms, the most commonly used mutation strategy calculation formula "DE / rand / 1" is:

[0140]

[0141] Among them, V i t represents the i-th mutation vector generated in the t-th iteration, where i∈[1,N V ]. r1,r2,r3∈[1,N V ] are three different integers. F i t is the i-th scaling factor in the t-th iteration, which controls the weight of the differential operation and generates the i-th mutation vector. To ensure the efficiency of the algorithm, its value is usually selected between 0 and 2. The new mutation vector matrix generated by the mutation strategy is:

[0142]

[0143] 3. Boundary processing method.

[0144] Vector boundary processing is usually performed after the mutation strategy. Due to the randomness of the test vector, the dimension of the newly generated parameter vector may exceed the feasible region, that is, exceed the preset boundary range of the linear phase-shifting transformer optimization design parameters. Therefore, it is necessary to perform boundary processing on the vector to ensure that all parameters in the optimization process always remain within the preset feasible region. The boundary processing method of the present invention is to pull the dimension that exceeds the boundary back to the boundary. Its mathematical expression is:

[0145]

[0146] Where t represents the tth iteration, i represents the i-th vector in the matrix, i∈[1,N V ], j represents the j-th dimension variable j∈[1,D] in a certain vector.

[0147] 4. Cross-strategy.

[0148] After boundary processing, the algorithm's crossover strategy is executed. The crossover strategy is to randomly replicate a certain dimension between different vectors X and crossover vectors V within certain rules to generate a new trial vector U.

[0149] The new trial vector stores information about both the old vector X and the new crossover vector V, and combines this information in a certain way to explore new potential ranges. Therefore, the crossover strategy enables the differential evolution algorithm to further enhance its global search capabilities within the feasible domain. The crossover strategy formula is:

[0150]

[0151] in, represents the j-th dimension variable of the i-th trial vector in the t-th iteration. CR∈[0,1] represents the crossover probability, which is a constant that needs to be defined manually. rand represents a random number uniformly distributed between 0 and 1, j rand is a random integer randomly selected from the range [1,2,…,D] to ensure that at least one dimension of the newly generated trial vector comes from the crossover vector. This effectively ensures that the vector evolves in a more optimal direction.

[0152] The trial vector matrix is ​​expressed as:

[0153]

[0154] For the optimization design of linear phase-shifting transformers, the significance of this strategy is actually to recombine different optimization design parameters to explore the impact of new design parameter combinations on the performance of linear phase-shifting transformers.

[0155] 5. Select a strategy.

[0156] The differential evolution algorithm uses the optimal criterion to select the next generation parameter vector X. The method is to evaluate the test vector U and the parameter vector X of the previous generation and select the vector with better fitness as the parameter vector of the next generation. The selection strategy is implemented as follows:

[0157]

[0158] Here, f[] represents the fitness value of the vector. In the linear phase-shifting transformer optimization design problem, the fitness value is actually the objective function value calculated by substituting the set of parameters into the equivalent circuit. Because the linear phase-shifting transformer optimization problem is to maximize the objective function, the optimal criterion should select the vector with the larger objective function value among the test vectors or parameter vectors as the next generation of parameter vectors. This ensures that the vectors in the matrix constantly move toward optimization as the number of iterations increases, thereby achieving global optimization more efficiently.

[0159] This embodiment of the present invention optimizes the design of a linear phase-shifting transformer based on an equivalent circuit model and an optimized design model, employing a differential evolution algorithm. Furthermore, this optimization method reduces approximations and assumptions in the design, reduces reliance on design experience, and conserves computing resources.

[0160] The linear phase-shifting transformer optimization design method provided by the embodiment of the present invention is divided into two levels: an outer algorithm layer and an inner analytical calculation layer. The algorithm layer uses its own learning method to continuously provide higher-quality parameter vectors to the internal analytical calculation layer. The inner calculation layer performs a detailed analysis of the parameter vector provided by the outer layer, ultimately calculating the objective function value of the parameter vector and providing feedback based on this value to the parameter vector provided by the outer algorithm layer. The algorithm layer adjusts its learning strategy based on this feedback and updates the parameter vector. The inner and outer layers work together to form a feedback channel, which continuously updates the parameters of the linear phase-shifting transformer and adjusts them towards a more optimal state.

[0161] The core structures of the primary and secondary sides of the linear phase-shifting transformer are completely symmetrical and fixed. The comparison results of the electromagnetic iterative design and the objective function-driven optimization design are shown in Table 2.

[0162] Table 2 Comparison results between electromagnetic iterative design and objective function driven optimization design

[0163] Design parameters unit Iterative design method optimization Primary and secondary side core width (2a) mm 108 100 <![CDATA[Slot width to slot pitch ratio (b s / τ s )]]> - 0.7 0.5 <![CDATA[Cross-sectional area of the wire (S con )]]> <![CDATA[mm 2 ]]> 3.38 1.06 Air gap length (g) mm 0.328 0.3 <![CDATA[Slot width (b s )]]> mm 7.8 6 <![CDATA[Slot height (h1)]]> mm 20 25 <![CDATA[Tooth width (t s )]]> mm 3.3 5 <![CDATA[Primary total current (I1)]]> A 8.3 6.05 <![CDATA[Current density (J c )]]> <![CDATA[A / m 3 ]]> 6 3 Single-side tooth weight (Weight) kg 5.99 5.22 Efficiency (η) ％ 82.25 94.36 Power Factor (PF) ％ 74.84 83.57 Voltage regulation (Δu) ％ 17.88 5.29 Harmonic content (THD) ％ 9.02 3.22

[0164] After optimizing the linear phase-shifting transformer using the optimization method provided by the present invention, the core weight is reduced by 10%, the efficiency is increased by 12.11%, the power factor is increased by 8.73%, the voltage regulation rate is reduced by 12.59, and the harmonic content is reduced by 5.8.

[0165] It should be noted that the embodiments of the present invention can be implemented by hardware, software, or a combination of software and hardware. The hardware portion can be implemented using dedicated logic; the software portion can be stored in a memory and executed by an appropriate instruction execution system, such as a microprocessor or dedicated design hardware. Those skilled in the art will appreciate that the above-mentioned devices and methods can be implemented using computer-executable instructions and / or contained in processor control code, for example, such as a carrier medium such as a disk, CD or DVD-ROM, a programmable memory such as a read-only memory (firmware), or a data carrier such as an optical or electronic signal carrier. The devices and modules of the present invention can be implemented by hardware circuits such as very large-scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, or programmable hardware devices such as field programmable gate arrays, programmable logic devices, etc., can also be implemented by software executed by various types of processors, or can be implemented by a combination of the above-mentioned hardware circuits and software, such as firmware.

[0166] The above description is only a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications, equivalent substitutions and improvements made by any technician familiar with this technical field within the technical scope disclosed by the present invention and within the spirit and principles of the present invention should be covered by the scope of protection of the present invention.

Claims

1. A linear phase-shifting transformer optimization design method, characterized in that: The linear phase-shifting transformer optimization design method includes: extracting several key design parameters as optimization variables based on the equivalent circuit of the linear phase-shifting transformer; defining the optimization design constraints and objective function applicable to the linear phase-shifting transformer model and establishing the optimization design model; using the differential evolution algorithm for parameter optimization design and using the equivalent circuit to optimize the linear phase-shifting transformer design; The optimization design using differential evolution algorithm includes: (1) Initialization settings In the initialization phase, the differential evolution algorithm randomly generates N V D-dimensional vectors, where NV represents the number of vectors in the matrix, which is set to 50. In one iteration, information interaction and learning are performed through 50 different solution vectors. D represents the number of variables contained in the vector, which refers to the number of design parameters contained in the vector, and is set to 5. In the initial solution matrix, the j-th dimension variable in the i-th vector is generated using the following formula: Among them, i represents the i-th vector in the matrix, i∈[1,N V ]; j represents the j-th dimension variable in the vector, j∈[1,D]; represents the j-th dimension variable of the i-th vector in the initial solution matrix, represents the j-th optimization variable of the i-th group of parameters; L(j) represents the lower boundary of the j-th dimension of the vector; U(j) represents the upper boundary of the j-th dimension of the vector; rand represents a random number uniformly distributed between 0 and 1; the generated initial solution matrix is: The initial solution matrix contains N V vectors, each with D variables; each vector represents a candidate solution, and the D variable values ​​in the candidate solution represent D design parameter values; in each iteration, information interaction and intelligent learning are achieved between different vectors, allowing each vector to evolve in a better direction; ultimately, the optimal vector is output, in which the D variables are the optimal values ​​of the design parameters of the linear phase-shifting transformer; (2) Mutation strategy Mutation strategy in N V Differential operations are performed between different vectors X to generate mutation vectors V. The order and position of the parameters contained in each vector remain unchanged during the optimization process, and only the values ​​change continuously through learning. For the differential evolution algorithm, the mutation strategy calculation formula "DE / rand / 1" is: Among them, V i t Represents the i-th mutation vector generated in the t-th iteration, i∈[1,N V ]; r1, r2, r3∈[1, N V ] are three different integers; F i t is the i-th scaling factor in the t-th iteration, which is used to control the weight of the differential operation and generates the i-th mutation vector, F i t The value is set between 0 and 2; The new mutation vector matrix generated by the mutation strategy is: (3) Boundary processing Pull the dimension beyond the boundary back to the boundary. The mathematical expression is: Where t represents the tth iteration; i represents the i-th vector in the matrix, i∈[1,N V ]; j represents the j-th dimension variable in a vector, j∈[1,D]; (4) Crossover strategy After the boundary processing, the algorithm crossover strategy is executed; the crossover strategy is to perform random replication of a certain dimension between different vectors X and crossover vectors V within the rule to generate a new trial vector U; The new trial vector stores information about both the old vector X and the new cross vector V, and combines the information in a certain way to explore new potential ranges. The crossover strategy enables the differential evolution algorithm to further enhance the global search capability within the feasible domain. The crossover strategy formula is: in, represents the j-th dimension variable of the i-th trial vector in the t-th iteration; CR∈[0,1] represents the crossover probability, which is a constant to be defined; rand represents a random number uniformly distributed between 0 and 1; j rand represents a random integer randomly selected from the range [1,2,…,D]; The trial vector matrix is ​​expressed as: For the optimization design of linear phase-shifting transformer, the crossover strategy is used to recombine different optimization design parameters to analyze the impact of the new design parameter combination on the performance of the linear phase-shifting transformer; (5) Select strategy The differential evolution algorithm uses the optimal criterion to select the next generation parameter vector X; the test vector U and the previous generation parameter vector X are evaluated, and the vector with better fitness is selected as the next generation parameter vector; The selection strategy is performed as follows: Among them, f[] represents the fitness value of the vector. In the linear phase-shifting transformer optimization design problem, the fitness value is the objective function value calculated by substituting the parameters into the equivalent circuit. Since the linear phase-shifting transformer optimization problem is to maximize the objective function, the optimal criterion selects the vector with a larger objective function value among the test vectors or parameter vectors as the next generation parameter vector.

2. The linear phase-shifting transformer optimization design method according to claim 1, characterized in that: The linear phase-shifting transformer optimization design method also includes: inputting algorithm parameters and initializing the parameter matrix; selecting the global optimal solution according to the objective function; judging whether Fes is less than Max_Fes and generating a matrix of test vectors; performing boundary processing and generating a matrix of cross vectors; selecting a better vector from the test vectors and the cross vectors, and finally selecting the global optimal solution according to the objective function value.

3. The linear phase-shifting transformer optimization design method according to claim 1, wherein: The linear phase-shifting transformer optimization design method includes the following steps: Step 1: Determine the optimized design model: Analyze the equivalent circuit model of the linear phase-shifting transformer and extract the parameters to be optimized in the equivalent circuit model; Step 2: Use the differential evolution algorithm for optimization design: Use the differential evolution algorithm to optimize the various parameters in the equivalent circuit model and finally obtain the optimal output result.

4. The linear phase-shifting transformer optimization design method according to claim 3, wherein: The determination of the optimal design model in step 1 includes: (1) Optimization parameter selection In the optimization design of a linear phase-shifting transformer, a total of five design parameters and / or variables are selected for optimization. The parameter optimization range is within a valid feasible domain, where the feasible domain is a value range. The selected design objects are arranged into groups in a fixed order, and each group of sequentially arranged design objects constitutes a vector. Based on the original design data of the linear phase-shifting transformer and combined with the parameters of the experimental prototype, the value range of the parameter objects is optimized. (2) Optimization design constraints Conditional constraints refer to the actual conditions that the linear phase-shifting transformer designed with parameters within the feasible region meets during operation. Based on references and the actual design of linear phase-shifting transformers, the following constraints are defined: B tooth ≤1.4T; a / τ≥0.7; R sf ≤0.7； gd≥0.2mm; Among them, B tooth Refers to the tooth magnetic flux amplitude, a is half of the primary core thickness, τ is the pole pitch, R sf represents the slot filling rate, g is the air gap length, and d is the secondary core thickness; The tooth flux density amplitude is not allowed to exceed 1.4T. The core stacking thickness to pole pitch ratio constrains the pole pitch size, preventing it from being too wide and reducing the length of the winding ends. In actual engineering, the slot fill factor limit is 0.75, and it is set to 0.7 to facilitate actual production and manufacturing. In addition, due to the small air gap of the linear induction linear phase-shifting transformer, the minimum air gap between the primary and secondary surfaces of the core is set to 0.2mm. (3) Determination of the objective function The slot fill rate is taken as the optimization target of the linear phase-shifting transformer. The efficiency, power factor, linear phase-shifting transformer body weight and slot fill rate are selected as multiple optimization targets to construct a comprehensive multi-objective optimization function to guide the optimization design process of the linear phase-shifting transformer. The constructed comprehensive multi-objective function expression is: Among them, η represents efficiency, PF is power factor, R sf is the slot fill rate, Weight is the tooth weight; k i Represents the weight coefficient of the optimization target, i=1,2,3.

5. The linear phase-shifting transformer optimization design method according to claim 4, characterized in that: The value range of the optimization parameter object in step (1) includes: The winding methods include full-pitch stacked winding on the primary side and long-short pitch winding on the secondary side; The minimum core stack thickness is 40mm and the maximum is 100mm; The minimum value of the slot width to slot pitch ratio is 0.3 and the maximum value is 0.7; The minimum tooth height is 10mm and the maximum is 50mm; The minimum current density is 3×10 6 A / mm 2 , the maximum value is 6×10 6 A / mm 2 ; The minimum value of the air gap length is 0.02mm and the maximum value is 0.05mm.

6. A linear phase-shifting transformer optimization design system using the linear phase-shifting transformer optimization design method according to any one of claims 1 to 5, characterized in that: The linear phase-shifting transformer optimization design system includes: An optimization design model building module is used to determine the optimization design model by optimizing parameter selection, optimizing design constraints, and determining the objective function; The transformer optimization design module is used to optimize the design of linear phase-shifting transformers using the differential evolution algorithm, including initialization settings, mutation strategy, boundary processing, crossover strategy and selection strategy.

7. A computer device, characterized in that: The computer device includes a memory and a processor. The memory stores a computer program. When the computer program is executed by the processor, the processor executes the steps of the linear phase-shifting transformer optimization design method according to any one of claims 1 to 5.

8. A computer-readable storage medium storing a computer program, wherein when the computer program is executed by a processor, the processor executes the steps of the linear phase-shifting transformer optimization design method according to any one of claims 1 to 5.

9. An information data processing terminal, characterized in that: The information data processing terminal is used to implement the linear phase-shifting transformer optimization design system as described in claim 6.

Images