A grounding net fault diagnosis method and system based on EVO-MNR algorithm

Abstract

The application relates to a grounding net fault diagnosis method and system based on an EVO-MNR algorithm, and the provided method comprises the following steps: acquiring all accessible nodes based on a grounding net topological structure, optionally injecting an excitation current source into the grounding net in turns by two accessible nodes, measuring and calculating the voltages of other accessible nodes to obtain measurement values and calculation values; taking the residual error of the measurement values and the calculation values as a target function to solve a first optimal solution of the grounding net branch resistance by using an EVO algorithm; acquiring a corresponding resistance value distribution based on the first optimal solution, solving corresponding calculation values based on the distribution; calculating a target function value based on the corresponding calculation values, if the calculated target function value is smaller than a preset value, the first optimal solution is an optimal solution, otherwise, the target function is improved to obtain an improved target function by using an MNR method, and the optimal solution of the grounding net branch resistance is solved based on the improved target function; and the fault branch is determined based on the optimal solution. Compared with the prior art, the application can quickly and accurately diagnose the grounding net fault.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of electrical equipment maintenance, in particular to a grounding grid fault diagnosis method and system based on an EVO-MNR algorithm. BACKGROUND

[0002] The existing maintenance method of the grounding grid of the transformer substation still stays in the stage of periodic large-area excavation inspection. This method not only takes time and effort, but also needs power-off maintenance, which inevitably brings economic losses. Moreover, in different soil environments, the corrosion rate of the grounding grid has obvious differences, and the soil environment is affected by many factors such as climate and geology, so it is difficult to determine the appropriate excavation maintenance period, which leads to the fact that some grounding grids have been severely corroded before excavation inspection, while some grounding grids are in good condition and do not need to be excavated. Therefore, this blind excavation and maintenance method not only cannot grasp the grounding grid condition in time and eliminate safety hazards, but also may cause waste of manpower and financial resources. Under the trend of increasing installed capacity and higher voltage level of the power system, the requirement for the performance of the grounding grid is also increasing, so it is of great significance to study an efficient and non-excavation grounding grid fault diagnosis technology to grasp the corrosion condition of the grounding grid in time. The performance of the grounding grid cannot meet the requirements due to the corrosion of the grounding grid material, which causes the failure of the transformer substation. Once the grounding grid of the transformer substation fails due to corrosion and fracture, it will endanger the safety and stability of the entire power system and may cause serious safety accidents. Therefore, it is of great practical significance to study a corrosion fault diagnosis technology for the grounding grid quickly and accurately without blind excavation. From the existing research, the current detection methods for the grounding grid of the transformer substation mainly include the electric network method, the magnetic field detection method and the electrical impedance imaging method, etc. For example, Chinese application patent CN113687191A combines the electrical impedance imaging and the electric network theory analysis method, uses the fault branch sensitivity characteristics of the grounding grid to quickly detect the fault of the grounding grid, and only needs to measure the port resistance without measuring the branch resistance and solving the equation set. The soil separation method is used for corrosion site imaging, although the process of measuring the branch resistance is avoided, but when the grounding grid fault is diagnosed, although only the port resistance is measured to simplify the workload, the data detected is complex, so the processing process is complicated, which leads to poor accuracy and has its own limitations, and the grounding grid of the transformer substation cannot be accurately diagnosed and analyzed.

[0003] Therefore, it is a technical problem to provide a method for quickly and accurately diagnosing the fault of the grounding grid. SUMMARY

[0004] The purpose of this invention is to overcome the shortcomings of the existing technology and provide a grounding grid fault diagnosis method and system based on the EVO-MNR algorithm. It combines the advantages of the electric network method in terms of fast and convenient location and the electrical impedance imaging method in terms of accuracy and intuitiveness. It proposes to apply the EVO-MNR algorithm to the grounding grid fault diagnosis solution process, which has the advantages of high solution accuracy and intuitive detection.

[0005] The objective of this invention can be achieved through the following technical solutions:

[0006] According to a first aspect of the present invention, a grounding grid fault diagnosis method based on the EVO-MNR algorithm is provided, comprising:

[0007] Based on the grounding grid topology, all accessible nodes are obtained. Two accessible nodes are randomly selected to inject excitation current sources into the grounding grid in turn. The voltages of other accessible nodes are measured and calculated to obtain the measured and calculated values.

[0008] The residual between the measured and calculated values ​​is used as the objective function to solve the first optimal solution of the grounding grid branch resistance using the EVO algorithm.

[0009] Based on the first optimal solution, obtain the corresponding resistance value distribution, and then calculate the corresponding value based on the distribution.

[0010] The objective function value is calculated based on the corresponding calculated value. If the calculated objective function value is less than the preset value, the first optimal solution is the optimal solution. Otherwise, the objective function is improved using the MNR method to obtain an improved objective function. The optimal solution of the grounding grid branch resistance is obtained based on the improved objective function.

[0011] The faulty branch is determined based on the aforementioned optimal solution.

[0012] As a preferred technical solution, the objective function expression is:

[0013]

[0014] Among them, U n (R) represents the calculated value of the branch resistance; This indicates the measured value of the branch resistance.

[0015] As a preferred technical solution, the method for solving the first optimal solution includes:

[0016] A feasible region is obtained for the resistance value distribution of each branch in the grounding grid topology. Multiple multidimensional resistance value distribution vectors are randomly generated in the feasible region as the initial particle population of the EVO algorithm.

[0017] The following steps are performed iteratively on the initial population:

[0018] Each particle population is evaluated based on the objective function to obtain the neutron enrichment level and enrichment boundary of the particle population; the position of each particle in the particle population is a candidate solution for the branch resistance.

[0019] The stability of the particles is evaluated based on the aforementioned objective function;

[0020] The positions of each particle in the particle population are updated based on the neutron enrichment level, enrichment boundary and stability. The particle population after position update is judged. If the iteration termination condition is met, the iteration ends and the current candidate solution is output as the first optimal solution.

[0021] If the iteration termination condition is not met, the next iteration will proceed based on the particle population updated with new positions.

[0022] As a preferred technical solution, the method for updating the position includes:

[0023] When the neutron enrichment level is greater than the enrichment boundary and the stability is greater than the objective function value of the particle with the best stability in the initial population, the position update of the particle population is performed according to α decay and γ decay; the expression for α decay is:

[0024]

[0025] in, X represents the new position of particle i under α decay; i Indicates the current position of particle i; X BS This represents the position vector of the particle with the best stability in the particle population; α represents the j-th component of the position vector of the i-th particle; II Indicates the index of the position component of alpha decay;

[0026] The expression for γ decay is:

[0027]

[0028] in, X represents the new position of particle i under γ decay; i Indicates the current position of particle i; X Ng This represents the position vector of the neighboring particles around the i-th particle; γ represents the j-th component of the position vector of the i-th particle; II Indicates the index of the position component of γ decay;

[0029] When the neutron enrichment level is greater than the enrichment boundary and the stability is less than or equal to the objective function value of the particle with the best stability in the initial population, the position update of the particle population is performed according to β decay; the expression for β decay is:

[0030]

[0031] in, and Both represent the new position of particle i under β decay; X i Indicates the current position of particle i; r1, r2, r3, and r4 are all random numbers; X BS X represents the position vector of the particle with optimal stability in the particle swarm; CP SL represents the position vector of a specific reference particle in the current population; i X represents the stability of the i-th particle; Ng This represents the position vector of the neighboring particles around the i-th particle;

[0032] When the neutron enrichment level is less than or equal to the enrichment boundary, the particle population position is updated randomly, and its expression is: in, X represents the new position of particle i. i Let r represent the current position of particle i, and r represent the random displacement variable.

[0033] As a preferred technical solution, the expression for the improved objective function is:

[0034]

[0035] in, U represents the introduced regularization penalty function; n (R) represents the calculated value of the branch resistance; R represents the measured value of the branch resistance; α represents the adjustable regularization coefficient; R0 represents the initial value of the branch resistance; R = [R1, R2, ..., R B ] T Let L represent the resistance value of each branch in B branches, and L represent the regularized identity matrix.

[0036] As a preferred technical solution, the method for obtaining the optimal solution based on the improved objective function is as follows:

[0037] If the calculated objective function value is greater than or equal to the preset value, then the improved objective function is Taylor expanded at the first optimal solution;

[0038] The iterative equation is obtained based on the Taylor expansion result;

[0039] The optimal solution is obtained based on the described iterative format.

[0040] As a preferred technical solution, the method for obtaining the iterative equation is as follows:

[0041] The improved objective function is applied to the first optimal solution R. (k) Performing a Taylor expansion at point gives the Taylor expansion, which is expressed as:

[0042]

[0043] The iterative format obtained based on the aforementioned Taylor expansion is expressed as follows: in,

[0044] The iterative formula is rearranged to obtain the iterative equation, whose expression is:

[0045]

[0046] Among them, R (k+1) R represents the branch resistance value in the (k+1)th iteration. (k) U represents the branch resistance value in the k-th iteration; n (R (k) U represents the calculated voltage corresponding to the branch resistance value in the k-th iteration; n0 Indicates the measured value; J k denoted as Jacobi matrix; L denotes the regularized identity matrix.

[0047] As a preferred technical solution, the method further includes: drawing a corrosion image of the faulty branch using electrical impedance imaging technology, wherein the method for drawing the corrosion image includes:

[0048] Compare the optimal solution with the nominal value of the branch resistance. If there is a deviation, it means that the branch is corroded.

[0049] Obtain the excitation current and field conductivity distribution of the corroded branch, and use the finite element method to construct a mathematical model of electric field distribution and boundary voltage based on the excitation current source and field conductivity distribution.

[0050] The mathematical model is linearized to obtain a linear relationship between the boundary voltage change and the conductivity distribution;

[0051] Construct an objective function for solving the inverse problem and an iterative update rule, and invert the resistivity in the field based on the objective function for solving the inverse problem, the iterative update rule and the linear relationship;

[0052] Corrosion images are drawn based on the inversion results.

[0053] As a preferred technical solution, the objective function expression for solving the inverse problem is:

[0054]

[0055] Where U(ρ) represents the calculated voltage value corresponding to the resistivity distribution ρ after the injection of the excitation current source; V represents the measured value of the boundary voltage; α1 represents the regularization coefficient; ρ 0 Indicates the initial resistivity;

[0056] The iterative update rule expression is as follows:

[0057] ρ (k+1) =ρ (k) -(J T J+α1L T L) -1 [J T (U(ρ)-V)+α1L T L(ρ (k) -ρ 0 )],

[0058] Where, ρ (k+1) ρ represents the result of the (k+1)th resistivity iteration; (k) denoted by ; J represents the partial derivative matrix of the calculated voltage with respect to resistivity; L represents the regularization identity matrix.

[0059] According to a second aspect of the present invention, a grounding grid fault diagnosis system based on the EVO-MNR algorithm is provided for implementing the method described herein.

[0060] Compared with the prior art, the present invention has the following advantages:

[0061] 1) This invention employs the EVO-MNR algorithm for fault diagnosis of grounding grids. The EVO algorithm searches for the optimal solution space of branch resistance globally based on the voltage measurement values ​​of accessible nodes in the grounding grid. By actively optimizing, it avoids the diagnostic blind spot caused by the assumption that the sensitivity of faulty branches is linearly separable in existing technologies. It is particularly suitable for the diagnostic needs of gradual degradation of multi-branch resistance in complex corrosion scenarios. Furthermore, based on this global optimization, a regularization parameter is introduced to improve the objective function of the EVO algorithm. The MNR method is used to reduce the dimensionality of multivariate data, effectively solving the problem of decreased computational efficiency caused by the surge in the complexity of solving equations when multiple faulty branches are coupled. The solution provided by this invention can achieve rapid and accurate diagnosis of grounding grid faults.

[0062] 2) Based on the EVO-MNR algorithm, this invention can further combine electrical impedance imaging technology to achieve visualization and precise location of the fault area. By constructing an electrical impedance distribution image, the resistivity distribution of each branch in the fault area can be displayed intuitively, thereby enabling rapid location of the fault and repair. Attached Figure Description

[0063] Figure 1 This is a flowchart of the method of the present invention;

[0064] Figure 2 This is a schematic diagram of the grounding grid topology in an embodiment of the present invention;

[0065] Figure 3 This is a flowchart of the electrical impedance imaging process of the present invention;

[0066] Figure 4 (4a) is a schematic diagram of the electrical impedance imaging results of the present invention; (4b) is a schematic diagram of grounding grid corrosion; (4c) is a corrosion image of the grounding grid containing soil; and (4d) is a corrosion image of the grounding grid after soil separation.

[0067] Figure 5 This is a schematic diagram of the software structure of Embodiment 3 of the present invention. Detailed Implementation

[0068] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0069] Unless otherwise defined, the technical or scientific terms used in this application shall have the ordinary meaning understood by one of ordinary skill in the art to which this application pertains. The terms “a,” “an,” “an,” “the,” and similar words used in this application do not indicate quantity limitation and may indicate singular or plural. The terms “comprising,” “including,” “having,” and any variations thereof used in this application are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or device that includes a series of steps or modules (units) is not limited to the listed steps or units, but may also include steps or units not listed, or may include other steps or units inherent to these processes, methods, products, or devices. The terms “connected,” “linked,” “coupled,” and similar words used in this application are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. “Multiple” used in this application refers to two or more. “And / or” describes the relationship between related objects, indicating that three relationships may exist; for example, “A and / or B” can represent: A alone, A and B simultaneously, and B alone. The character " / " generally indicates that the preceding and following objects are in an "or" relationship. The terms "first," "second," and "third" used in this application are merely to distinguish similar objects and do not represent a specific ordering of the objects.

[0070] Example 1

[0071] This invention addresses the shortcomings of existing technologies by applying an intelligent algorithm (EVO-MNR algorithm) to the grounding grid fault diagnosis process, combined with a combined electrical network method, aiming to achieve rapid and efficient fault diagnosis of grounding grids. The detailed method flow provided by this invention is as follows: Figure 1 As shown, it includes:

[0072] S1. Obtain the measured and calculated values ​​of the accessible node voltages under different excitation current sources.

[0073] S11. In detail, refer to the substation grounding grid construction drawings to obtain the grounding grid topology. In this embodiment, obtain a simple grounding grid topology containing nine reachable nodes, such as... Figure 2 As shown.

[0074] S12. Based on the grounding grid topology, obtain all accessible nodes, randomly select two accessible nodes to inject excitation current sources into the grounding grid in turn, measure and calculate the voltage of other accessible nodes, and obtain the measured value and the calculated value.

[0075] S2. Using the residual between the measured and calculated values ​​as the objective function, the EVO algorithm is used to find the first optimal solution for the grounding grid branch resistance.

[0076] S21. Objective function setting.

[0077] The EVO algorithm is an intelligent optimization algorithm inspired by particle decay, which seeks optimization by simulating particle collision decay and the mechanism of migration to a steady state. In grounding grid diagnosis, this algorithm takes the branch resistance vector R as the optimization objective, that is, the position of each particle in the particle population is a candidate solution for the branch resistance. Faults are located by minimizing the residual between the calculated and measured values ​​of the node voltage. In specific implementation, an initial population is first generated in the feasible resistance region, and the decay mode is dynamically adjusted according to the fitness function, etc. The particles update the resistance parameters according to the rules, and the optimization is iterated until convergence. The optimal resistance solution is output to determine the conductor corrosion state, realizing efficient inversion of parameters of complex resistance networks. Its goal in grounding grid diagnosis is to find R such that the residual between the calculated and measured values ​​of the node voltage under different excitation modes is minimized. Specifically, the objective function expression is:

[0078]

[0079] Among them, U n (R) represents the calculated value of the branch resistance; This indicates the measured value of the branch resistance.

[0080] S22. Obtain the feasible region of the resistance value distribution of each branch in the grounding grid topology, and randomly generate n resistance value distribution vectors with dimension d in the feasible region as the initial particle population of the EVO algorithm.

[0081] The details are as follows:

[0082]

[0083] Where X represents the total number of particles in the search space; X n This represents the nth particle; This represents the j-th decision variable of the i-th particle, i.e., the resistance value of the j-th branch; and The lower and upper limits of the resistance value of the j-th branch are represented; rand represents a random number uniformly distributed in the range [0,1].

[0084] The following steps are performed iteratively on the initial population:

[0085] S23. Evaluate each particle population based on the objective function to obtain the neutron enrichment level (NEL) and enrichment boundary (EB) of the particle population. The calculation expression is as follows:

[0086]

[0087] Among them, NEL i This represents the neutron enrichment level of the i-th particle.

[0088] S24. The stability of particles is evaluated based on the objective function, the expression of which is:

[0089]

[0090] Where BS represents the maximum value of the objective function in the initial population; WS represents the minimum value of the objective function in the initial population; SL i This represents the stability of the i-th particle.

[0091] S25. Update the positions of each particle in the particle population based on neutron enrichment level, enrichment boundary, and stability. Evaluate the updated particle population; if the iteration termination condition is met, end the iteration and output the current candidate solution X. BS As the first optimal solution.

[0092] Detailed methods for updating the location include:

[0093] i. When the neutron enrichment level is greater than the enrichment boundary and the stability is greater than the objective function value of the particle with the best stability in the initial population, the position update of the particle population is carried out according to α decay and γ decay.

[0094] α decay: α is randomly generated in the range [1, d]. I In [1,α I Randomly generate α within the range II Its expression is:

[0095]

[0096] in, X represents the new position of particle i under α decay; i Indicates the current position of particle i; X BS This represents the position vector of the particle with the best stability in the particle population; α represents the j-th component of the position vector of the i-th particle; II Indicates the index of the position component of alpha decay;

[0097] γ decay: γ is randomly generated in the range [1, d]. I , in [1,γ I Randomly generate γ within the range II Its expression is:

[0098]

[0099] in, X represents the new position of particle i under γ decay; i Indicates the current position of particle i; X Ng This represents the position vector of the neighboring particles around the i-th particle; γ represents the j-th component of the position vector of the i-th particle; II Indicates the index of the position component of γ decay;

[0100] ii. When the neutron enrichment level is greater than the enrichment boundary and the stability is less than or equal to the objective function value of the particle with the best stability in the initial population, the position update of the particle population is performed according to β decay.

[0101] The expression for β decay is:

[0102]

[0103] in, and Both represent the new position of particle i under β decay; X i Indicates the current position of particle i; r1, r2, r3, and r4 are all random numbers; X BS X represents the position vector of the particle with optimal stability in the particle swarm; CP SL represents the position vector of a specific reference particle in the current population; i X represents the stability of the i-th particle; NgThis represents the position vector of the neighboring particles around the i-th particle;

[0104] iii. When the particle enrichment level is less than or equal to the enrichment boundary, the position update method of the particle population is random update, and its expression is: in, X represents the new position of particle i. i Let r represent the current position of particle i, and r represent the random displacement variable.

[0105] S26. If the iteration termination condition is not met, then execute steps S23 to S25 again based on the particle population after position update.

[0106] S3. Obtain the corresponding resistance value distribution based on the first optimal solution, and calculate the corresponding value based on the distribution.

[0107] S4. Calculate the objective function value based on the corresponding calculated value. If the calculated objective function value is less than the preset value, then the first optimal solution is the optimal solution. Otherwise, improve the objective function using the MNR method to obtain the improved objective function. Based on the improved objective function, find the optimal solution for the grounding grid branch resistance.

[0108] The Newton-Raphson (NR) algorithm is an efficient mathematical method for solving nonlinear algebraic problems. It can linearize the nonlinear equation of the objective function in R... k After performing a Taylor expansion at the extreme point, and substituting the first and second derivatives of the function f1 into the equation, we can obtain the iterative solution as follows:

[0109]

[0110] In the formula, J k J is the Jacobian matrix, representing the partial derivative of the node voltage with respect to the branch resistance. k The condition number is relatively large, which may lead to The larger condition number leads to ill-conditioned problems in the solution process. Therefore, a regularization coefficient objective function is introduced to improve the solution, resulting in an improved objective function, the expression of which is:

[0111]

[0112] in, U represents the introduced regularization penalty function; n (R) represents the calculated value of the branch resistance; R represents the measured value of the branch resistance; α represents the adjustable regularization coefficient; R0 represents the initial value of the branch resistance; R = [R1, R2, ..., R B ] T Let L represent the resistance value of each branch in B branches, and L represent the regularized identity matrix.

[0113] By solving the optimal problem corresponding to the improved objective function, approximate resistance values ​​of the grounding grid branch conductors can be obtained. By comparing these resistance values ​​with the nominal resistance values ​​of each branch, the corrosion status of the conductors in each branch can be determined. The specific solution process is as follows:

[0114] S41. If the improved objective function is applied to the first optimal solution R (k) Performing a Taylor expansion at point gives the Taylor expansion, which is expressed as:

[0115]

[0116] S42. Take the partial derivative of the Taylor expansion, and find the extrema based on the obtained partial derivatives. The expression is as follows:

[0117]

[0118] Based on the above formula, the iterative format is obtained, and its expression is as follows:

[0119]

[0120] S43. Rearrange the iterative format to obtain the iterative equation, the expression of which is:

[0121]

[0122]

[0123] Among them, R (k+1) R represents the branch resistance value in the (k+1)th iteration. (k) U represents the branch resistance value in the k-th iteration; n (R (k) U represents the calculated voltage corresponding to the branch resistance value in the k-th iteration; n0 Indicates the measured value; J k denoted as Jacobi matrix; L denotes the regularized identity matrix.

[0124] Through the above regularization process, The problem of finding the inverse is transformed into the above formula. The problem of matrix inversion can be improved by changing α to adjust its eigenvalues, thereby improving the condition number of ill-conditioned matrices and thus the problem of matrix inversion. (Details follow.) The calculation steps for (the Jacobian matrix under the excitation current source in the e-th term) are as follows:

[0125]

[0126] in, This represents a highly sparse matrix, where only the element in the j-th row and j-th column is 0. The rest are 0.

[0127] S43. Find the optimal solution based on the iterative format.

[0128] S5. Based on the optimal solution, determine the faulty branch and use electrical impedance imaging technology to draw corrosion images of the faulty branch. The process flow is as follows: Figure 3 As shown.

[0129] S51. Compare the optimal solution with the nominal value of the branch resistance. If there is a deviation, it means that the branch is corroded.

[0130] S52. Obtain the excitation current and field conductivity distribution of the corroded branch, and use the finite element method to construct a mathematical model of the electric field distribution and boundary voltage based on the excitation current source and field conductivity distribution.

[0131] The mathematical model is as follows:

[0132]

[0133] Where ΔU is the change in the measured boundary voltage; Δφ is the distribution of the conductivity change.

[0134] S53. To simplify subsequent calculations, we now simplify the mathematical model to obtain a linear relationship between boundary voltage change and conductivity distribution, i.e., b = Ag, where b = ΔU, g = Δφ, and A represents the sensitivity matrix.

[0135] S54. Construct the objective function for solving the inverse problem and the iterative update rule, and invert the resistivity in the field based on the objective function for solving the inverse problem, the iterative update rule and the linear relationship.

[0136] S541. Construct the inverse problem and solve the objective function.

[0137] Because the inverse problem suffers from severe ill-conditioning, this invention employs a Gauss-Newton iterative method based on Tikhonov regularization to solve the electrical impedance imaging problem. Its expression is:

[0138]

[0139] Where U(ρ) represents the calculated voltage value of the resistivity distribution ρ of the corresponding N partitioned triangular elements after the injection of the excitation current source; V represents the measured value of the boundary voltage; α1 represents the regularization coefficient; ρ 0 This represents the initial resistivity.

[0140] S542. Construct iterative update rules.

[0141] The update rule for the resistivity of multiple triangulated units in each iteration is as follows:

[0142]

[0143] Where, ρ (k+1) ρ represents the result of the (k+1)th resistivity iteration; (k) denoted by ; J represents the partial derivative matrix of the calculated voltage with respect to resistivity; L represents the regularization identity matrix.

[0144] S543, Inverted resistivity.

[0145] Finally, an image of the grounding grid's impedance distribution is reconstructed. By measuring the potential or current signals on the object's surface and using image reconstruction algorithms, the inverse problem can be solved to inversely calculate the internal conductivity or resistivity distribution of the object. This provides crucial information about the object's internal structure and function, facilitating non-invasive monitoring and diagnosis. Although the inverse problem solution process is complex and prone to ill-posedness, techniques such as regularization can effectively improve the accuracy and stability of the reconstructed image.

[0146] S55. Draw a corrosion image based on the inversion results.

[0147] In this embodiment, a corrosion image as shown in Figure (4a) is constructed. Corrosion images with and without soil are constructed as shown in Figures (4b) and (4c), respectively. Comparing Figures (4a), (4b) and (4c), it can be seen that the method provided by the present invention is feasible and has high diagnostic accuracy.

[0148] Example 2

[0149] As a preferred technical solution, the EVO algorithm is used to optimize the regularization coefficient α and the branch resistance value, so that the regularization coefficient can be adaptively adjusted according to the needs. This can effectively prevent overfitting and underfitting while ensuring the performance of the algorithm used in the invention. By optimizing the regularization coefficient through the EVO algorithm, it is no longer necessary to set it manually, which can ensure that the results are more accurate and objective when performing diagnosis.

[0150] In detail, the steps of this embodiment include:

[0151] A1. Based on the obtained grounding grid topology, obtain the measured and calculated values ​​of the accessible node voltages under different excitation current sources.

[0152] A2. Using the residual between the measured and calculated values ​​as the objective function, the EVO algorithm is used to find the first optimal solution for the grounding grid branch resistance.

[0153] A3. Obtain the corresponding resistance value distribution based on the first optimal solution, and solve for the corresponding calculated value based on the distribution.

[0154] A4. Calculate the objective function value based on the corresponding calculated value. If the calculated objective function value is less than the preset value, then the first optimal solution is the optimal solution. Otherwise, improve the objective function using the MNR method to obtain the improved objective function. Use the EVO algorithm to use the improved objective function as the fitness function to optimize the regularization coefficient α. Substitute the optimized α into the improved objective function to obtain the optimal solution of the grounding grid branch resistance.

[0155] A5. Based on the optimal solution, determine the faulty branch and use electrical impedance imaging technology to draw corrosion images of the faulty branch.

[0156] For detailed operations of each step in this embodiment, please refer to the relevant content in Embodiment 1, which will not be repeated here.

[0157] Example 3

[0158] This embodiment provides a software system for grounding grid fault diagnosis. This software system is a human-computer interactive grounding grid fault joint diagnosis software design platform. Through a graphical interface, users can operate intuitively without directly manipulating code, simplifying the process and improving work efficiency. The software system includes six modules: "User Management," "Parameter Configuration," "Information Acquisition," "EVO-MNR," "Impact Imaging," and "Result Analysis." Figure 5 As shown. In the development of grounding grid fault diagnosis software, the joint operation and optimization debugging of various modules are key to ensuring software performance and stability. After preliminary requirements analysis and system design, as well as the research and development of fault diagnosis algorithms, the various functional modules have been initially formed. At this point, they need to be seamlessly integrated to form a collaborative whole. First, the data acquisition module, fault diagnosis algorithm module, and result display module are assembled according to the design architecture to ensure that data can flow smoothly between the modules. By simulating actual grounding grid fault scenarios, inputting test data, and observing the interaction between the modules, as well as the overall software response speed and diagnostic accuracy, we need to pay special attention to the interface compatibility and data transmission efficiency between modules to ensure that information is not distorted or lost during transmission. Problems found during joint operation are investigated and resolved one by one. For the algorithm module, parameters are adjusted according to the test results, and the algorithm logic is optimized to improve the accuracy and efficiency of diagnosis. For the data acquisition module, sensor configuration and data acquisition strategies are optimized to reduce noise interference and improve data quality. The result display module is beautified to make it more intuitive and easy to use, so that users can quickly understand the diagnostic results.

[0159] Its usage process is as follows:

[0160] (1) Log in to the interface based on the user's identity. If you do not have an account, you need to register a personal account to log in to the software system.

[0161] (2) Users input substation grounding grid information according to the actual situation of the grounding grid. The software system builds a grounding grid model based on the grounding grid information input by the user, including two parts: basic grounding grid parameters and topology parameters. The basic grounding grid parameters include grounding grid conductor resistivity, grounding grid conductor size, soil resistivity, etc. The topology parameters mainly include grounding grid node information, branch information, and combinations of accessible nodes, etc.

[0162] (3) Import the excitation mode and node voltage acquisition data of the external current source, and import the node voltage information under different excitation modes acquired in the hardware device into the system.

[0163] (4) Fault diagnosis: The software system processes the above data using the EVO-MNR algorithm. The output result is a bar chart of the calculated resistance values ​​of each branch of the grounding grid. If it is a single branch fault, the result can be directly output and jump to the next step. If it is a multi-branch fault, the impedance imaging of the fault area where the fault branch is located can be further selected.

[0164] (5) Perform impedance imaging on the fault area to obtain the resistivity distribution map of the fault area, which makes it easier for users to intuitively judge the degree of corrosion fault of the grounding grid.

[0165] (6) Output the diagnostic results in tabular form, including images and diagnostic conclusions during the diagnostic process.

[0166] During the optimization and debugging of the software system, an iterative development approach is adopted, involving continuous testing, feedback, and modification until the software performance reaches its optimal level. Furthermore, it is crucial to emphasize the software's robustness and fault tolerance, ensuring stable operation even under extreme conditions and providing users with reliable grounding grid fault diagnosis services.

[0167] Furthermore, this embodiment also provides a grounding grid fault diagnosis system based on the EVO-MNR algorithm as the hardware system corresponding to the aforementioned software system. This hardware system includes a central processing unit (CPU), which can execute various appropriate actions and processes based on computer program instructions stored in read-only memory (ROM) or loaded from storage units into random access memory (RAM). The RAM can also store various programs and data required for device operation. The CPU, ROM, and RAM are interconnected via a bus. Input / output (I / O) interfaces are also connected to the bus.

[0168] Multiple components in the device are connected to the I / O interface, including: input units such as keyboards and mice; output units such as various types of displays and speakers; storage units such as disks and optical discs; and communication units such as network interface cards (NICs), modems, and wireless transceivers. The communication unit allows the device to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0169] The processing unit executes the various methods and processes described above, such as methods S1-S5 and A1-A5. For example, in some embodiments, methods S1-S5 and A1-A5 may be implemented as computer software programs tangibly contained in a machine-readable medium, such as a storage unit. In some embodiments, part or all of the computer program may be loaded and / or installed on the device via ROM and / or a communication unit. When the computer program is loaded into RAM and executed by the CPU, one or more steps of methods S1-S5 and A1-A5 described above may be performed. Alternatively, in other embodiments, the CPU may be configured to execute methods S1-S5 and A1-A5 by any other suitable means (e.g., by means of firmware).

[0170] The functions described above in this document can be performed, at least in part, by one or more hardware logic components. For example, exemplary types of hardware logic components that can be used, without limitation, include: Field Programmable Gate Arrays (FPGAs), Application-Specific Integrated Circuits (ASICs), Application Standard Products (ASSPs), System-on-Chip (SoCs), Complex Programmable Logic Devices (CPLDs), and so on.

[0171] The program code used to implement the methods of the present invention can be written in any combination of one or more programming languages. This program code can be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing device, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code can be executed entirely on the machine, partially on the machine, as a standalone software package partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0172] In the context of this invention, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. Machine-readable media can include, but are not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0173] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in the present invention, and these modifications or substitutions should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A grounding grid fault diagnosis method based on the EVO-MNR algorithm, characterized in that, include: Based on the grounding grid topology, all accessible nodes are obtained. Two accessible nodes are randomly selected to inject excitation current sources into the grounding grid in turn. The voltages of other accessible nodes are measured and calculated to obtain the measured and calculated values. The residual between the measured and calculated values ​​is used as the objective function to solve the first optimal solution of the grounding grid branch resistance using the EVO algorithm. Based on the first optimal solution, obtain the corresponding resistance value distribution, and then calculate the corresponding value based on the distribution. The objective function value is calculated based on the corresponding calculated value. If the calculated objective function value is less than the preset value, the first optimal solution is the optimal solution. Otherwise, the objective function is improved using the MNR method to obtain an improved objective function. The optimal solution of the grounding grid branch resistance is obtained based on the improved objective function. The faulty branch is determined based on the aforementioned optimal solution.

2. The grounding grid fault diagnosis method based on the EVO-MNR algorithm according to claim 1, characterized in that, The objective function expression is: ， in, This represents the calculated value of the branch resistance; This indicates the measured value of the branch resistance.

3. The grounding grid fault diagnosis method based on the EVO-MNR algorithm according to claim 1, characterized in that, The methods for finding the first optimal solution include: A feasible region is obtained for the resistance value distribution of each branch in the grounding grid topology. Multiple multidimensional resistance value distribution vectors are randomly generated in the feasible region as the initial particle population of the EVO algorithm. The following steps are performed iteratively on the initial population: Each particle population is evaluated based on the objective function to obtain the neutron enrichment level and enrichment boundary of the particle population; the position of each particle in the particle population is a candidate solution for the branch resistance. The stability of the particles is evaluated based on the aforementioned objective function; The positions of each particle in the particle population are updated based on the neutron enrichment level, enrichment boundary and stability. The particle population after position update is judged. If the iteration termination condition is met, the iteration ends and the current candidate solution is output as the first optimal solution. If the iteration termination condition is not met, the next iteration will proceed based on the particle population updated with new positions.

4. The grounding grid fault diagnosis method based on the EVO-MNR algorithm according to claim 3, characterized in that, The methods for updating the location include: When the neutron enrichment level is greater than the enrichment boundary and the stability is greater than the objective function value of the particle with the best stability in the initial population, the particle population position is updated according to... decay and Decay is used for updating; The decay expression is: ， in, express The new position of particle i after decay; Indicates the current position of particle i; This represents the position vector of the particle with the best stability in the particle population; This represents the j-th component of the position vector of the i-th particle; express The decay location component index; The decay expression is: in, express The new position of particle i after decay; Indicates the current position of particle i; This represents the position vector of the neighboring particles around the i-th particle; This represents the j-th component of the position vector of the i-th particle; express The decay location component index; When the neutron enrichment level is greater than the enrichment boundary and the stability is less than or equal to the objective function value of the particle with the best stability in the initial population, the particle population position is updated according to... The decay process is updated; the aforementioned The decay expression is: ， ， in, and All indicate The new position of particle i after decay; Indicates the current position of particle i; All represent random numbers; This represents the position vector of the particle with the best stability in the particle population; This represents the position vector of a specific reference particle in the current population; This represents the stability of the i-th particle; This represents the position vector of the neighboring particles around the i-th particle; When the neutron enrichment level is less than or equal to the enrichment boundary, the particle population position is updated randomly, and its expression is: ,in, Indicates the new position of particle i. This indicates the current position of particle i. This represents a random displacement variable.

5. The grounding grid fault diagnosis method based on the EVO-MNR algorithm according to claim 1, characterized in that, The expression for the improved objective function is as follows: ， in, This represents the introduced regularization penalty function; This represents the calculated value of the branch resistance; The measured value of the branch resistance; Indicates the adjustable regularization coefficient; Indicates the initial value of the branch resistance; This represents the resistance value of each branch in branch B; Represents the regularized identity matrix; This represents the objective function.

6. The grounding grid fault diagnosis method based on the EVO-MNR algorithm according to claim 1, characterized in that, The method for obtaining the optimal solution based on the improved objective function is as follows: If the calculated objective function value is greater than or equal to the preset value, then the improved objective function is Taylor expanded at the first optimal solution; The iterative equation is obtained based on the Taylor expansion result; The optimal solution is obtained based on the described iterative format.

7. A grounding grid fault diagnosis method based on the EVO-MNR algorithm according to claim 6, characterized in that, The method for obtaining the iterative equation is as follows: The improved objective function is applied to the first optimal solution. Performing a Taylor expansion at point gives the Taylor expansion, which is expressed as: ； The iterative format obtained based on the aforementioned Taylor expansion is expressed as follows: ;in, , ; The iterative formula is rearranged to obtain the iterative equation, whose expression is: ， in, Indicates the first The branch resistance value in the next iteration; Indicates the first The branch resistance value in the next iteration; Indicates the first The calculated voltage value corresponding to the branch resistance value in the next iteration; Indicates the measured value; Represents the Jacobian matrix; Represents the regularized identity matrix; Indicates the adjustable regularization coefficient; Indicates the initial value of the branch resistance; This represents the resistance value of each branch in branch B.

8. The grounding grid fault diagnosis method based on the EVO-MNR algorithm according to claim 1, characterized in that, The method further includes: using electrical impedance imaging technology to draw a corrosion image of the faulty branch, wherein the method for drawing the corrosion image includes: Compare the optimal solution with the nominal value of the branch resistance. If there is a deviation, it means that the branch is corroded. Obtain the excitation current and field conductivity distribution of the corroded branch, and use the finite element method to construct a mathematical model of electric field distribution and boundary voltage based on the excitation current source and field conductivity distribution. The mathematical model is linearized to obtain a linear relationship between the boundary voltage change and the conductivity distribution; Construct an objective function for solving the inverse problem and an iterative update rule, and invert the resistivity in the field based on the objective function for solving the inverse problem, the iterative update rule and the linear relationship; Corrosion images are drawn based on the inversion results.

9. A grounding grid fault diagnosis method based on the EVO-MNR algorithm according to claim 8, characterized in that, The objective function expression for solving the inverse problem is: ， in, This represents the resistivity distribution after the injection of the excitation current source. Calculated voltage value; The measured value of the boundary voltage; Represents the regularization coefficient; Indicates the initial resistivity; The iterative update rule expression is as follows: ， in, Indicates the first The result of the resistivity iteration; This represents the result of the k-th resistivity iteration; The matrix representing the partial derivatives of the calculated voltage with respect to resistivity; This represents the regularized identity matrix.

10. A grounding grid fault diagnosis system based on the EVO-MNR algorithm, characterized in that, The system is used to implement the method as described in any one of claims 1 to 9.

Images