Rapid inversion method for high-frequency equivalent circuit parameters of transformer

Abstract

A rapid inversion method for high-frequency equivalent circuit parameters of a transformer comprises the steps that a universal high-frequency equivalent circuit model suitable for windings of any layer number is built based on a modular series topological structure, a time domain differential equation set is built through the Kirchhoff voltage law, and circuit parameters are initialized; the time domain equation is converted into a complex frequency domain algebraic equation set, and theoretical frequency response is calculated in combination with power parameters; the method comprises the following steps: calibrating a pole position by using FRA experimental data, quantifying a distance relationship between an extreme point and a measuring point of a frequency response curve through a multi-Gaussian function dynamic weighting algorithm, and constructing a weighted deviation objective function; and adopting a dimension traversal type simulated annealing algorithm, realizing global optimization of high-dimensional parameters through candidate solution generation, adaptive step length adjustment, progressive temperature attenuation and a double termination judgment mechanism, and outputting an equivalent circuit parameter set. The method breaks through the limitation of a traditional model on the number of winding layers, remarkably improves the parameter identification precision and efficiency, and provides high-reliability technical support for transformer health state evaluation and early fault diagnosis.

Description

Technical Field

[0001] The present invention relates to transformer state monitoring, and in particular to a fast inversion method for high-frequency equivalent circuit parameters of a transformer. Background Art

[0002] Transformers, as core electrical equipment in power systems, have a direct impact on the safety of the power grid through their operational status. Accurately assessing transformer health and providing timely warnings of early faults are crucial for improving power system reliability. Frequency response analysis (FRA), with its sensitive detection of subtle mechanical defects such as winding deformation, turn-to-turn shorts, and loose clamps, has become a promising non-invasive diagnostic method for transformer condition monitoring.

[0003] In the application of FRA technology, equivalent circuit parameter inversion is a key technical bottleneck that needs to be broken through urgently, which is mainly reflected in three dimensions: (1) At the modeling theory level, the existing simulation methods are limited by commercial software platforms and lack a universal high-frequency equivalent circuit mathematical model suitable for windings with any number of layers, which makes it difficult to implement programmed analysis and parameter optimization; (2) At the feature extraction level, the existing inversion methods do not have enough information mining on the extreme points of the frequency response curve, making it difficult to effectively capture the morphological evolution law of the frequency response curve; (3) At the algorithm implementation level, the parameter identification process involves dozens of variables to be optimized (the specific dimensions depend on the circuit topology, usually not less than 10 dimensions), and traditional optimization algorithms based on gradient information are prone to fall into local optimality, affecting the accuracy of parameter identification.

[0004] It should be noted that the information disclosed in the above background technology section is only used to understand the background of this application, and therefore may include information that does not constitute prior art known to ordinary technicians in this field. Summary of the Invention

[0005] The main purpose of the present invention is to overcome the defects existing in the above-mentioned background technology and provide a method for fast inversion of high-frequency equivalent circuit parameters of a transformer.

[0006] To achieve the above object, the present invention adopts the following technical solutions:

[0007] A method for quickly inverting high-frequency equivalent circuit parameters of a transformer comprises the following steps:

[0008] S1. Import transformer frequency response analysis (FRA) experimental data and construct a universal high-frequency equivalent circuit model for windings with any number of layers. This model characterizes the distributed parameters of the windings through a modular series topology. Based on Kirchhoff's voltage law, a set of time-domain differential equations is established to initialize the equivalent circuit parameters.

[0009] S2. Converting the time domain differential equations into a complex frequency domain algebraic equations, and calculating the theoretical frequency response of the equivalent circuit by combining the voltage amplitude and angular frequency parameters of the external power supply;

[0010] S3. Based on the FRA experimental data, the pole position is calibrated, the weight value of each measuring point is calculated by the multi-Gaussian function dynamic weighting algorithm, and the objective function with the weight value and the theoretical-experimental deviation as the core is constructed;

[0011] S4. A dimensional traversal simulated annealing algorithm is used to solve the minimum value of the objective function through a fourth-order optimization framework of candidate solution generation, adaptive step size control, progressive temperature decay and double termination judgment, and output an equivalent circuit parameter set.

[0012] Furthermore, step S1 specifically includes:

[0013] Defining circuit parameters corresponding to each winding layer based on a modular series topology, wherein the parameters include mutual inductance between windings, capacitance between layers, and distributed resistance;

[0014] The correlation equations of the node voltage and branch current of the high-voltage winding and the low-voltage winding are established by hierarchical analytical time-domain differential equations, and the constraints of the external excitation source are introduced.

[0015] Furthermore, step S2 specifically includes:

[0016] Perform Laplace transform on the time domain differential equations to generate a complex frequency domain algebraic equation system containing complex frequency variables;

[0017] The frequency response function is defined based on the measurement connection method, and the amplitude-frequency characteristics of the theoretical frequency response are obtained by simultaneously solving the complex frequency domain equations.

[0018] Furthermore, step S3 specifically includes:

[0019] The extreme points of the measured frequency response curve are extracted as feature constraints, and a multi-Gaussian combination weight function is constructed based on the distance between the extreme point position and the frequency measurement point;

[0020] The mean of the weighted sum of squares of theoretical-experimental deviations is used as the objective function to quantify the matching degree between the frequency response curve morphology and parameter inversion.

[0021] Furthermore, the dimension traversal simulated annealing algorithm in step S4 specifically includes:

[0022] (1) Candidate solution generation: Perform a single-dimensional random perturbation on the current solution in dimensional order to generate candidate solutions and calculate the objective function value;

[0023] (2) Acceptance probability judgment: Based on the current temperature and the objective function difference between the candidate solution and the current solution, the probability of accepting the inferior solution is calculated;

[0024] (3) Adaptive step size control: Dynamically adjust the search step size based on the acceptance rate of candidate solutions to balance global exploration and local convergence capabilities;

[0025] (4) Progressive temperature decay: Lower the temperature value in stages, and gradually reduce the probability of accepting inferior solutions through an exponential cooling strategy;

[0026] (5) Double termination judgment: The optimization termination condition is determined by jointly determining the preset iteration number threshold and the continuous optimization amplitude threshold.

[0027] Furthermore, the adaptive step size control specifically includes:

[0028] The step size range is adjusted in real time according to the acceptance probability of the candidate solution. When the acceptance probability exceeds the upper limit, the step size is increased to speed up the search. When the acceptance probability is lower than the lower limit, the step size is reduced to improve the accuracy.

[0029] Furthermore, the progressive temperature attenuation specifically includes:

[0030] After completing the traversal search of all dimensions in each temperature cycle, the temperature value is reduced according to a fixed cooling coefficient, so that the algorithm gradually transitions from global random search to local fine optimization.

[0031] Furthermore, the double termination determination specifically includes:

[0032] Monitor the optimization range of the objective function over multiple consecutive temperature cycles. If the optimization range continues to be lower than the preset threshold and has not reached the maximum number of iterations, it is determined to be convergence termination; otherwise, forced termination is performed.

[0033] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the method for fast inversion of high-frequency equivalent circuit parameters of a transformer.

[0034] A computer program product includes a computer program, wherein when the computer program is executed by a processor, the method for fast inversion of high-frequency equivalent circuit parameters of a transformer is implemented.

[0035] The present invention has the following beneficial effects:

[0036] The present invention addresses the technical bottleneck in the inversion of high-frequency equivalent circuit parameters of transformers and proposes a solution with comprehensive innovative advantages. The fast inversion method for high-frequency equivalent circuit parameters of transformers of the present invention breaks through the restriction of traditional methods to a specific number of winding layers by constructing a universal high-frequency equivalent circuit model of modular series topology structure, realizes a unified mathematical representation of transformers with any number of winding layers, and solves the problems of programmed analysis and cross-model adaptation caused by model incompatibility. In terms of frequency response feature analysis, a multi-Gaussian dynamic weighted algorithm is innovatively introduced. By establishing a nonlinear mapping relationship between the extreme point distance and the weight coefficient, the morphological evolution law of the frequency response curve is quantitatively characterized, and the frequency response feature capture capability of subtle defects such as winding deformation is significantly improved. In response to the challenge of easily falling into local optimality in high-dimensional parameter optimization, a dimensional traversal simulated annealing algorithm framework is designed, which integrates candidate solution generation, adaptive step size control, progressive temperature decay and dual termination judgment mechanism, effectively balancing global search and convergence accuracy, and overcoming the defect of premature convergence of traditional gradient algorithms in high-dimensional space. This method combines the synergistic advantages of universal modeling, precise feature analysis and efficient parameter optimization, providing more accurate non-invasive diagnostic technology support for transformer health status assessment and fault warning.

[0037] Other beneficial effects of the embodiments of the present invention will be further described below. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 This is an overall flow chart of the method for rapid inversion of high-frequency equivalent circuit parameters of a transformer according to the present invention.

[0039] Figure 2 This is an algorithm flow chart of a preferred embodiment of the transformer high-frequency equivalent circuit parameter inversion method of the present invention.

[0040] Figure 3 This is a topological diagram of the high-frequency equivalent circuit of the transformer according to an embodiment of the present invention.

[0041] Figure 4 Schematic diagram of the extreme point position calibration of the transformer FRA curve according to an embodiment of the present invention.

[0042] Figure 5 This is a flow chart of the simulated annealing method for optimizing equivalent circuit parameters according to an embodiment of the present invention. DETAILED DESCRIPTION

[0043] The following is a detailed description of the embodiments of the present invention. It should be emphasized that the following description is only exemplary and is not intended to limit the scope of the present invention and its application.

[0044] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be understood to indicate or imply relative importance or implicitly specify the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of such features. In the description of the embodiments of the present invention, "plurality" means two or more, unless otherwise specifically defined.

[0045] See Figure 1 The embodiment of the present invention provides a method for fast inversion of high-frequency equivalent circuit parameters of a transformer, comprising the following steps:

[0046] Step S1. Import transformer frequency response analysis (FRA) experimental data and construct a universal high-frequency equivalent circuit model for windings with any number of layers. The model characterizes the distributed parameters of the windings through a modular series topology structure, establishes a time-domain differential equation group based on Kirchhoff's voltage law, and initializes the equivalent circuit parameters.

[0047] In some embodiments, step S1 specifically includes: defining the circuit parameters corresponding to each winding layer based on a modular series topology, wherein the parameters include the mutual inductance between windings, the interlayer capacitance and the distributed resistance; establishing the correlation equations of the node voltage and branch current of the high-voltage winding and the low-voltage winding through layered analytical time-domain differential equations, and introducing the constraints of the external excitation source.

[0048] Step S2: converting the time domain differential equations into a complex frequency domain algebraic equations, and calculating the theoretical frequency response of the equivalent circuit by combining the voltage amplitude and angular frequency parameters of the external power supply.

[0049] In some embodiments, step S2 specifically includes: performing Laplace transform on the time domain differential equation group to generate a complex frequency domain algebraic equation group containing complex frequency variables; defining a frequency response function based on the measurement wiring method, and obtaining the amplitude-frequency characteristics of the theoretical frequency response by simultaneously solving the complex frequency domain equation group.

[0050] Step S3. Based on the FRA experimental data (such as the measured frequency response curve), the pole position is calibrated, the weight value of each measuring point is calculated by the multi-Gaussian function dynamic weighting algorithm, and the objective function with the weight value and the theoretical-experimental deviation as the core is constructed.

[0051] In some embodiments, step S3 specifically includes: extracting the extreme points of the measured frequency response curve as feature constraints, constructing a multi-Gaussian combination weight function based on the distance between the extreme point position and the frequency measurement point; using the mean of the weighted sum of squares of theoretical-experimental deviations as the objective function, and quantifying the matching degree between the frequency response curve morphology and parameter inversion.

[0052] Step S4. Using a dimensional traversal simulated annealing algorithm, through a fourth-order optimization framework of candidate solution generation, adaptive step size control, progressive temperature decay and double termination judgment, the minimum value of the objective function is solved and an equivalent circuit parameter set is output.

[0053] In some embodiments, the dimension traversal simulated annealing algorithm in step S4 specifically includes (see Figure 5 ): (1) Candidate solution generation: Perform a single-dimensional random perturbation on the current solution in dimensional order to generate candidate solutions and calculate the objective function value; (2) Acceptance probability judgment: Calculate the probability of accepting inferior solutions based on the current temperature and the objective function difference between the candidate solution and the current solution; (3) Adaptive step size control: Dynamically adjust the search step size according to the candidate solution acceptance rate to balance global exploration and local convergence capabilities; (4) Progressive temperature decay: Reduce the temperature value in stages and gradually reduce the probability of accepting inferior solutions through an exponential cooling strategy; (5) Double termination judgment: The optimization termination condition is jointly determined by the preset iteration number threshold and the continuous optimization amplitude threshold.

[0054] In some embodiments, the adaptive step size control specifically includes: adjusting the step size range in real time according to the acceptance probability of the candidate solution, expanding the step size to accelerate the search when the acceptance probability exceeds the upper limit, and reducing the step size to improve the accuracy when the acceptance probability is lower than the lower limit.

[0055] In some embodiments, the progressive temperature attenuation specifically includes: after completing the traversal search of all dimensions in each temperature cycle, reducing the temperature value according to a fixed cooling coefficient, so that the algorithm gradually transitions from global random search to local fine optimization.

[0056] In some embodiments, the dual termination determination specifically includes: monitoring the optimization range of the objective function within multiple consecutive temperature cycles; if the optimization range continues to be lower than a preset threshold and does not reach the maximum number of iterations, it is determined to be convergence termination; otherwise, forced termination is performed.

[0057] The fast inversion method for transformer high-frequency equivalent circuit parameters of the present invention, by constructing a universal high-frequency equivalent circuit model applicable to winding structures with any number of layers (abbreviated as N layers), breaks through the technical bottleneck of traditional methods being limited to a specific number of winding layers, and adopts a modular series topology structure to realize a unified mathematical representation of transformer distributed parameters; in response to the defect of insufficient analysis of frequency response curve characteristics, an innovative dynamic weighting algorithm based on extreme value characteristic constraints is proposed, and a nonlinear mapping relationship between weight coefficients and the distance between extreme points of the frequency response curve is established through a multi-Gaussian function combination model, accurately quantifying the morphological evolution law of the frequency response curve; further combined with the simulated annealing algorithm framework, a dimensional traversal global optimization strategy is designed, integrating candidate solution generation, adaptive step size control, progressive temperature decay and dual termination judgment mechanism, effectively solving the problem that traditional gradient algorithms in high-dimensional parameter space are prone to falling into local optimal solutions, and significantly improving the global convergence and computational efficiency of parameter identification. The present invention significantly improves the accuracy and efficiency of parameter identification, providing high-reliability technical support for transformer health status assessment and early fault diagnosis.

[0058] The following further describes specific embodiments of the present invention and experimental verification.

[0059] A method for inverting high-frequency equivalent circuit parameters of a transformer, the algorithm flow of its preferred embodiment is as follows Figure 2 This method can be divided into three main parts: transformer high-frequency equivalent circuit modeling, high-frequency equivalent circuit frequency response calculation, and high-frequency equivalent circuit parameter inversion. These three parts will be described in detail below.

[0060] (1) Transformer high-frequency equivalent circuit modeling:

[0061] The topology of the transformer high-frequency equivalent circuit is as follows: Figure 3 As shown in Figure 1, the circuit topology consists of n circuit modules connected in series. Each module can simulate a single-disc winding or a multi-disc winding. The circuit components of each module have specific physical meanings, as shown in Table 1.

[0062] Table 1 Physical meaning of transformer equivalent circuit parameters

[0063]

[0064]

[0065] (2) Calculation of frequency response of transformer high-frequency equivalent circuit:

[0066] Figure 3 The frequency response of the circuit topology shown can be solved using Kirchhoff's Voltage Law (KVL), which contains Circuit parameters, including 14n basic circuit parameters (resistance, conductance, inductance, capacitance), n 2The high-voltage-low-voltage winding mutual inductance parameter, The mutual inductance parameters of the high voltage winding are: There are (4n+3) quantities to be solved in the circuit, including (n+1) high-voltage winding node voltages. (n+1) low voltage winding node voltages n high-voltage winding branch currents n low-voltage winding branch currents 1 external current I in The specific KVL equation is as follows (see Table 1 for parameter meanings):

[0067] The KVL time domain differential equation of the top winding is:

[0068]

[0069]

[0070] The KVL time-domain differential equation of the intermediate winding (i=1, 2, 3, ..., n-1) circuit is:

[0071] The KVL time domain differential equation of the bottom winding is:

[0072]

[0073]

[0074] Supplementary time domain equations:

[0075]

[0076] Performing Laplace transform on the differential equations shown in formulas (2), (3), and (4) yields the following algebraic equation for the top winding KVL in the complex frequency domain (s domain):

[0077]

[0078] The KVL complex frequency domain (s domain) algebraic equation for the intermediate winding (i = 1, 2, 3, ..., n-1) circuit is:

[0079]

[0080]

[0081] The KVL complex frequency domain (s domain) algebraic equation of the bottom winding is:

[0082]

[0083] Supplementary complex frequency domain (s domain) algebraic equations:

[0084] V0(s)-V n (s)=U in (s) (8)

[0085] The meanings of the parameters in formulas (1)-(8) can be obtained from Table 1. In addition, U in is the voltage amplitude of the external power supply. Substituting the specific s parameter (s=jω, ω is the angular frequency) into the simultaneous equations (5)-(8) can obtain (4n+3) linear algebraic equations, and (4n+3) quantities to be determined can be obtained. The frequency response of the transformer high-frequency equivalent circuit is related to the wiring method of the measuring device. Figure 3 Taking the most commonly used wiring method as an example, its frequency response is:

[0086]

[0087] Where f is the frequency of the external power supply.

[0088] (3) Inversion of transformer high-frequency equivalent circuit parameters:

[0089] Since the extreme position of the FRA curve contains rich transformer structural information, when performing parameter inversion of the transformer high-frequency equivalent circuit, it is necessary to focus on the measurement points near the extreme position. In order to achieve the above goals, the present invention adopts a strategy of combining multiple Gaussian functions for dynamic weight allocation. First, the extreme position of the FRA curve is calibrated. Taking the FRA curve of a certain transformer as an example, the extreme position calibration is as follows: Figure 4 It should be noted that not all extreme points of the FRA curve need to be calibrated, but only the extreme points that affect the overall trend of the FRA curve need to be calibrated.

[0090] Assume that a certain FRA curve has a total of m poles calibrated, and the jth pole parameter determined by observing the FRA curve is μ = μ j ,σ=σ j , then the weight value of the i-th measuring point on the FRA curve is calculated as follows:

[0091]

[0092] Wherein, xi is the frequency of the i-th measurement point. The objective function of the transformer high-frequency equivalent circuit parameter inversion adopted by the present invention is as follows:

[0093]

[0094] Where N is the number of sampling points of the FRA curve, x k is the kth sampling point of the FRA curve, x′ k is the calculated value of the kth sampling point, w(x k) is the weight value of the kth sampling point. At this point, the inversion problem of the transformer high-frequency equivalent circuit parameters is transformed into an optimization problem of solving the minimum value of the objective function shown in formula (11).

[0095] The present invention adopts simulated annealing method to solve the optimization problem. The simulated annealing method flow chart is as follows: Figure 5 As shown, where fx c and fx are the objective function values of the candidate solution and the current solution respectively. The core steps of the simulated annealing method include:

[0096] Candidate solution generation strategy:

[0097] Each iteration generates candidate solutions through the dimension traversal strategy: numerical perturbations are performed on only one dimension at a time, and the new value of the dimension is composed of the original value superimposed on the random coefficient ran∈(-1,1) and the product of the current step size. The values of the remaining dimensions remain constant, and the objective function value needs to be calculated after each dimension is adjusted. The step size update judgment is triggered only after all dimensions are completely traversed, so each step size cycle needs to be executed Second objective function evaluation.

[0098] Adaptive step size control:

[0099] Dynamic step size adjustment is based on the acceptance rate balance principle: when the acceptance probability of a dimension is too high, the step size is increased; when the acceptance rate is too low, the step size is reduced. This mechanism effectively balances search efficiency and accuracy, avoiding missing potential optimization areas due to excessively large step sizes, or local convergence due to excessively small step sizes.

[0100] Progressive temperature decay:

[0101] Phased cooling strategy: Temperature decay is performed after completing several step updates. T ∈(0,1) cooling coefficient (the present invention sets R T =0.8), causing the temperature to decay exponentially. As the temperature decreases, the probability of the algorithm accepting an inferior solution gradually decreases. This design ensures that the solution space is fully explored in the early stages and that stable convergence can be achieved in the later stages.

[0102] Double termination judgment:

[0103] A two-tier termination mechanism is implemented: 1) Forced termination: Immediate termination occurs when the total number of iterations exceeds a preset upper limit; 2) Convergence termination: Monitoring the last d temperature update cycles (preferably, d = 5) and determining convergence if the objective function optimization amplitude consistently falls below a sensitivity threshold. Both termination conditions are verified after the temperature update phase to ensure the output is the current optimal solution.

[0104] In summary, the present invention proposes a universal method for inverting high-frequency equivalent circuit parameters of transformers with any number of winding layers. Compared with existing methods for inverting high-frequency equivalent circuit parameters of transformers, the embodiments of the present invention have at least the following advantages:

[0105] 1. Breakthrough in the universal modeling framework:

[0106] A universal high-frequency equivalent circuit model for any number of winding layers is constructed, enabling unified characterization of transformers with any number of winding layers through a modular series topology. This overcomes the limitations of existing software platforms and establishes a programmable mathematical model for analysis. This addresses the model incompatibility issues inherent in traditional methods due to differences in the number of winding layers, enabling parameter optimization to be adaptable across transformer models.

[0107] 2. Improved frequency response feature analysis capabilities:

[0108] An innovative multi-Gaussian dynamic weighting algorithm is proposed to establish a nonlinear mapping relationship between extreme point distance and weight coefficient. Through the extreme value feature constraint mechanism, the morphological evolution of the frequency response curve is quantitatively characterized. This technology enables the parameter inversion process to accurately capture the subtle frequency response characteristics of winding deformation.

[0109] 3. Improved performance of high-dimensional parameter optimization:

[0110] A dimensional traversal simulated annealing algorithm is designed to solve the premature convergence problem of traditional gradient algorithms in high-dimensional parameter spaces through a four-order optimization framework of "candidate solution generation-adaptive step size-gradual cooling-double termination".

[0111] An embodiment of the present invention further provides a storage medium for storing a computer program, which at least performs the above method when executed.

[0112] An embodiment of the present invention further provides a control device, comprising a processor and a storage medium for storing a computer program; wherein the processor is configured to execute at least the method described above when executing the computer program.

[0113] An embodiment of the present invention further provides a processor, which executes a computer program and at least performs the method described above.

[0114] The storage medium can be implemented by any type of non-volatile storage device, or a combination thereof. Among them, the non-volatile memory can be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), a magnetic random access memory (FRAM), a flash memory (Flash Memory), a magnetic surface memory, an optical disc or a read-only optical disc (CD-ROM); the magnetic surface memory can be a disk memory or a tape memory. The storage medium described in the embodiments of the present invention is intended to include, but is not limited to, these and any other suitable types of memory.

[0115] In the several embodiments provided by the present invention, it should be understood that the disclosed systems and methods can be implemented in other ways. The device embodiments described above are merely illustrative. For example, the division of the units is merely a logical function division. In actual implementation, there may be other division methods, such as: multiple units or components can be combined, or can be integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the components shown or discussed can be through some interfaces, and the indirect coupling or communication connection of the devices or units can be electrical, mechanical or other forms.

[0116] The units described above as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units; some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0117] In addition, all functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may be separately used as a unit, or two or more units may be integrated into one unit; the above-mentioned integrated units may be implemented in the form of hardware or in the form of hardware plus software functional units.

[0118] Those skilled in the art will understand that all or part of the steps of the above-mentioned method embodiment can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps of the above-mentioned method embodiment; and the aforementioned storage medium includes: mobile storage devices, read-only memory (ROM), random access memory (RAM), magnetic disk or optical disk, etc. Various media that can store program codes.

[0119] Alternatively, if the above-mentioned integrated unit of the present invention is implemented in the form of a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the embodiment of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes a number of instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the methods described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as mobile storage devices, ROM, RAM, magnetic disks or optical disks.

[0120] The methods disclosed in the several method embodiments provided by the present invention can be arbitrarily combined without conflict to obtain new method embodiments.

[0121] The features disclosed in several product embodiments provided by the present invention can be arbitrarily combined without conflict to obtain new product embodiments.

[0122] The features disclosed in several method or device embodiments provided by the present invention can be arbitrarily combined without conflict to obtain new method embodiments or device embodiments.

[0123] The above is a further detailed description of the present invention in conjunction with specific preferred embodiments, and the specific implementation of the present invention should not be considered to be limited to these descriptions. Those skilled in the art will recognize that, without departing from the scope of the present invention, several equivalent substitutions or obvious variations can be made, and the performance or use of the same should be considered to fall within the scope of protection of the present invention.

Claims

1. A method for fast inversion of transformer high-frequency equivalent circuit parameters, characterized in that: The following steps are involved: S1. Import transformer frequency response analysis (FRA) experimental data and construct a universal high-frequency equivalent circuit model for windings with any number of layers. This model characterizes the distributed parameters of the windings through a modular series topology. Based on Kirchhoff's voltage law, a set of time-domain differential equations is established to initialize the equivalent circuit parameters. S2. Converting the time domain differential equations into a complex frequency domain algebraic equations, and calculating the theoretical frequency response of the equivalent circuit by combining the voltage amplitude and angular frequency parameters of the external power supply; S3. Based on the FRA experimental data, the pole position is calibrated, the weight value of each measuring point is calculated by the multi-Gaussian function dynamic weighting algorithm, and the objective function with the weight value and the theoretical-experimental deviation as the core is constructed; S4. A dimensional traversal simulated annealing algorithm is used to solve the minimum value of the objective function through a fourth-order optimization framework of candidate solution generation, adaptive step size control, progressive temperature decay and double termination judgment, and output an equivalent circuit parameter set.

2. The method for rapid inversion of high-frequency equivalent circuit parameters of a transformer according to claim 1, wherein: Step S1 specifically includes: Defining circuit parameters corresponding to each winding layer based on a modular series topology, wherein the parameters include mutual inductance between windings, capacitance between layers, and distributed resistance; The correlation equations of the node voltage and branch current of the high-voltage winding and the low-voltage winding are established by hierarchical analytical time-domain differential equations, and the constraints of the external excitation source are introduced.

3. The method for rapid inversion of transformer high-frequency equivalent circuit parameters according to claim 1, characterized in that: Step S2 specifically includes: Perform Laplace transform on the time domain differential equations to generate a complex frequency domain algebraic equation system containing complex frequency variables; The frequency response function is defined based on the measurement connection method, and the amplitude-frequency characteristics of the theoretical frequency response are obtained by simultaneously solving the complex frequency domain equations.

4. The method for rapid inversion of high-frequency equivalent circuit parameters of a transformer according to claim 1, wherein: Step S3 specifically includes: The extreme points of the measured frequency response curve are extracted as feature constraints, and a multi-Gaussian combination weight function is constructed based on the distance between the extreme point position and the frequency measurement point; The mean of the weighted sum of squares of theoretical-experimental deviations is used as the objective function to quantify the matching degree between the frequency response curve morphology and parameter inversion.

5. The method for rapid inversion of transformer high-frequency equivalent circuit parameters according to claim 1, wherein: The dimension traversal simulated annealing algorithm in step S4 specifically includes: (1) Candidate solution generation: Perform a single-dimensional random perturbation on the current solution in dimensional order to generate candidate solutions and calculate the objective function value; (2) Acceptance probability judgment: Based on the current temperature and the objective function difference between the candidate solution and the current solution, the probability of accepting the inferior solution is calculated; (3) Adaptive step size control: Dynamically adjust the search step size based on the acceptance rate of candidate solutions to balance global exploration and local convergence capabilities; (4) Progressive temperature decay: Lower the temperature value in stages, and gradually reduce the probability of accepting inferior solutions through an exponential cooling strategy; (5) Double termination judgment: The optimization termination condition is determined by jointly determining the preset iteration number threshold and the continuous optimization amplitude threshold.

6. The method for rapid inversion of transformer high-frequency equivalent circuit parameters according to claim 5, characterized in that: The adaptive step size control specifically includes: The step size range is adjusted in real time according to the acceptance probability of the candidate solution. When the acceptance probability exceeds the upper limit, the step size is increased to speed up the search. When the acceptance probability is lower than the lower limit, the step size is reduced to improve the accuracy.

7. The method for rapid inversion of high-frequency equivalent circuit parameters of a transformer according to claim 5, characterized in that: The progressive temperature attenuation specifically includes: After completing the traversal search of all dimensions in each temperature cycle, the temperature value is reduced according to a fixed cooling coefficient, so that the algorithm gradually transitions from global random search to local fine optimization.

8. The method for rapid inversion of transformer high-frequency equivalent circuit parameters according to claim 5, characterized in that: The double termination judgment specifically includes: Monitor the optimization range of the objective function over multiple consecutive temperature cycles. If the optimization range continues to be lower than the preset threshold and has not reached the maximum number of iterations, it is determined to be convergence termination; otherwise, forced termination is performed.

9. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the method for fast inversion of high-frequency equivalent circuit parameters of a transformer according to any one of claims 1 to 8 is implemented.

10. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the method for fast inversion of high-frequency equivalent circuit parameters of a transformer according to any one of claims 1 to 8 is implemented.

Images