Primary and secondary fusion ring network box electric field distribution optimization method and device

Abstract

The application provides a secondary fusion ring network box electric field distribution optimization method and device, relates to the technical field of electric field measurement, and comprises the following steps: obtaining ring network box design parameter groups of a target deployment position, and establishing a ring network box finite element model; according to the target deployment position, performing scene parameter feature mining under multiple working conditions; solving the electric field distribution anomaly of the ring network box finite element model under virtual mapping; performing multi-parameter sensitivity coupling correlation calculation on the ring network box design parameter groups; feeding back and adjusting the ring network box design parameter groups according to a ring network box design optimization guide map; and performing multi-working-condition collaborative electric field distribution risk optimization on the ring network box design adjustment space according to a plurality of working condition characteristic matrices. Through the application, multi-working-condition electric field characteristics and design parameter coupling modeling and collaborative analysis are realized, precise electric field anomaly identification and parameter optimization linkage are realized, and the technical effects of improving the insulation design optimization precision and the operation safety and stability are achieved.

Description

Technical Field

[0001] This application relates to the field of electric field measurement technology, and in particular to a method and apparatus for optimizing the electric field distribution of a primary and secondary integrated ring network box. Background Technology

[0002] As power systems continue to develop towards higher voltage levels, larger capacity power supply, and intelligent power distribution, the integrated primary and secondary ring main units, as key node equipment in urban power distribution networks, are gradually exhibiting characteristics such as multiple operating conditions superimposed, high electric field stress density, and complex electromagnetic environment coupling. During long-term operation, the equipment not only needs to withstand the steady-state voltage of the power frequency, but also needs to cope with the transient or local abnormal electrical stress effects such as lightning impulse overvoltage and partial discharge. Under the influence of multi-physical field coupling, the electric field distribution state inside the ring main unit exhibits significant spatial non-uniformity and time-varying fluctuations.

[0003] Currently, existing design and analysis methods for ring mainframes mainly rely on electromagnetic field simulation analysis or empirical design verification under single operating conditions. They typically use steady-state power frequency conditions as the primary design basis, supplemented by local impact checks for verification. During the design process, finite element analysis is often performed with fixed structural and material parameters, lacking a systematic modeling and unified evaluation mechanism for the differences in electric field response under different operating conditions. Furthermore, there is a lack of quantitative analysis methods for the nonlinear coupling relationship between design parameters and electric field anomalies, making it difficult to accurately identify localized areas of concentrated electric field and potential insulation weaknesses under multiple overlapping operating conditions. In addition, existing methods often use single-index or local threshold judgment methods in risk assessment, failing to conduct collaborative analysis of multiple sources such as power frequency voltage, lightning impulses, and partial discharges. They also cannot reflect the comprehensive impact of changes in design parameters on the overall structure of the electric field distribution, making the design optimization process still reliant on empirical adjustments or local trial calculations, lacking a global optimization guidance mechanism.

[0004] In summary, the existing technology suffers from a lack of a unified modeling and collaborative analysis mechanism for the coupling relationship between the electric field distribution characteristics under multiple operating conditions and design parameters. This makes it difficult to accurately identify and quantify abnormal electric field regions in ring main units under complex operating environments, further affecting the scientific nature of parameter adjustments and the reliability of overall electrical insulation performance during design optimization. Consequently, this restricts the safe and stable operation of ring main units under complex electromagnetic environments under multiple operating conditions. Summary of the Invention

[0005] The purpose of this application is to provide a method and apparatus for optimizing the electric field distribution of a primary and secondary integrated ring main unit (RMU). This aims to address the technical problem in the prior art where the lack of a unified modeling and collaborative analysis mechanism for the coupling relationship between the electric field distribution characteristics under multiple operating conditions and design parameters makes it difficult to accurately identify and quantify abnormal electric field regions in complex operating environments. This further affects the scientific nature of parameter adjustments and the reliability of overall electrical insulation performance during the design optimization process, thereby restricting the safe and stable operation of the RMU in complex electromagnetic environments under multiple operating conditions.

[0006] In view of the above problems, this application provides a method and apparatus for optimizing the electric field distribution of a primary and secondary integrated ring network box.

[0007] Firstly, this application provides a method for optimizing the electric field distribution of a primary and secondary integrated ring main unit (RMU), implemented using a primary and secondary integrated RMU electric field distribution optimization device. The method includes: obtaining a set of design parameters for the RMU at a target deployment location, and establishing a finite element model of the RMU based on the set of design parameters; performing scene parameter feature mining under multiple operating conditions based on the target deployment location to determine multiple operating condition characteristic matrices; solving for electric field distribution anomalies under virtual mapping on the RMU finite element model based on the multiple operating condition characteristic matrices to obtain a multi-operating condition electric field anomaly map; performing multi-parameter sensitivity coupling correlation calculation on the RMU design parameter set based on the multi-operating condition electric field anomaly map to establish a RMU design optimization guidance network; performing feedback adjustment on the RMU design parameter set based on the RMU design optimization guidance network to obtain the RMU design adjustment space; and performing multi-operating condition collaborative electric field distribution risk optimization on the RMU design adjustment space based on the multiple operating condition characteristic matrices to obtain the RMU design optimization result.

[0008] Preferably, the method for optimizing the electric field distribution of the primary and secondary integrated ring network box further includes: retrieving the power frequency voltage condition log set, lightning impulse condition log set, and partial discharge condition log set at the target deployment location; performing multi-parameter coupling relationship mining on the power frequency voltage condition log set to obtain a power frequency voltage scenario parameter association set; performing key parameter filtering and aggregation on the power frequency voltage condition log set based on the power frequency voltage scenario parameter association set to obtain a power frequency voltage condition characteristic matrix; performing multi-parameter coupling relationship mining and key parameter filtering and aggregation on the lightning impulse condition log set to obtain a lightning impulse condition characteristic matrix; performing multi-parameter coupling relationship mining and key parameter filtering and aggregation on the partial discharge condition log set to obtain a partial discharge condition characteristic matrix; and combining the power frequency voltage condition characteristic matrix and the lightning impulse condition characteristic matrix to generate the multiple condition characteristic matrices.

[0009] Preferably, the primary and secondary fusion ring network box electric field distribution optimization method further includes: mapping the multiple operating condition characteristic matrices to the ring network box finite element model respectively to obtain multiple virtual operating condition mapping models; performing mesh adaptive partitioning and boundary condition loading on the ring network box finite element model according to each virtual operating condition mapping model to obtain multiple operating condition electric field solution models; performing steady-state electric field identification, transient impact electric field identification, and local distortion electric field identification on each operating condition electric field solution model to obtain multiple operating condition electric field distribution results; extracting local maximum field strength parameters, electric field gradient abrupt change parameters, insulation interface field strength parameters, and field strength concentration region parameters according to each operating condition electric field distribution result to obtain multiple electric field anomaly feature sets; and performing anomaly region labeling and map encoding according to the multiple electric field anomaly feature sets to obtain the multi-operating condition electric field anomaly map.

[0010] Preferably, the primary and secondary fusion ring main unit electric field distribution optimization method further includes: extracting an electric field anomaly region set based on the multi-condition electric field anomaly map; determining multiple design parameter anomaly association pairs based on the positional correspondence and structural mapping relationship between the electric field anomaly region set and the ring main unit design parameter group; calculating the sensitivity value of each design parameter in the ring main unit design parameter group to local maximum field strength, electric field uniformity, insulation interface field strength, and partial discharge risk based on the multiple design parameter anomaly association pairs, and obtaining multiple design parameter sensitivity vectors; performing parameter coupling correlation analysis and cross-condition response consistency analysis on the ring main unit design parameter group based on the multiple design parameter sensitivity vectors, and obtaining a design parameter coupling correlation matrix; and performing graph structure encoding based on the multiple design parameter sensitivity vectors and the design parameter coupling correlation matrix to generate the ring main unit design optimization guidance graph network.

[0011] Preferably, the primary and secondary fusion ring network box electric field distribution optimization method further includes: extracting the Kth design parameter according to the ring network box design parameter group, where K is a positive integer; extracting the target abnormal region and target abnormal index corresponding to the Kth design parameter according to the multiple design parameter anomaly correlation pairs; subjecting the Kth design parameter to a preset amplitude perturbation while keeping other parameters in the ring network box design parameter group unchanged, and obtaining multiple ring network box design perturbation groups; mapping the multiple ring network box design perturbation groups to the ring network box finite element model respectively, and solving the electric field distribution for the target abnormal region to obtain multiple perturbation electric field results; analyzing the multiple perturbation electric field results to obtain the electric field change sequence corresponding to the target abnormal index, wherein the electric field change sequence includes the local maximum field strength change, electric field uniformity change, insulation interface field strength change, and partial discharge risk change; calculating the multi-index sensitivity value corresponding to the Kth design parameter according to the electric field change sequence, and normalizing and vectorizing the multi-index sensitivity value to obtain the Kth design parameter sensitivity vector.

[0012] Preferably, the primary and secondary fusion ring main unit electric field distribution optimization method further includes: filtering the ring main unit design adjustment space according to the ring main unit design constraints at the target deployment location to obtain a ring main unit design candidate space; performing electric field distribution risk optimization on the ring main unit design candidate space according to the power frequency voltage condition characteristic matrix to obtain a first optimal ring main unit design domain; performing electric field distribution risk optimization on the first optimal ring main unit design domain according to the lightning impulse condition characteristic matrix to obtain a second optimal ring main unit design domain; performing electric field distribution risk optimization on the second optimal ring main unit design domain according to the partial discharge condition characteristic matrix to obtain a third optimal ring main unit design domain; and performing minimum electric field non-uniformity risk optimization on the third optimal ring main unit design domain to generate the ring main unit design optimization result.

[0013] Preferably, the primary and secondary fusion ring network box electric field distribution optimization method further includes: fitting the electric field distribution of each ring network box design candidate scheme in the ring network box design candidate space according to the power frequency voltage condition characteristic matrix to obtain multiple power frequency voltage electric field distribution characteristics; training an electric field distribution risk prediction model based on the ring network box electric field distribution risk event set; inputting the multiple power frequency voltage electric field distribution characteristics into the electric field distribution risk prediction model to obtain multiple electric field distribution risk values; and filtering the ring network box design candidate space according to the electric field distribution risk threshold based on the multiple electric field distribution risk values ​​to obtain the first optimal ring network box design domain.

[0014] Preferably, the method for optimizing the electric field distribution of the primary and secondary integrated ring network box further includes: the ring network box design parameter group includes the ring network box structural parameter group, the ring network box material parameter group, and the ring network box electrical parameter group.

[0015] Preferably, the method for optimizing the electric field distribution of the primary and secondary integrated ring network box further includes: the set of abnormal electric field regions includes a region of local maximum electric field strength anomaly, a region of abrupt change in electric field gradient, a region of abnormal insulation interface, and a region of partial discharge risk.

[0016] Secondly, this application also provides a primary and secondary fusion ring network box electric field distribution optimization device for executing the primary and secondary fusion ring network box electric field distribution optimization method as described in the first aspect, including: a ring network box finite element model establishment module, used to obtain the ring network box design parameter set at the target deployment location and establish a ring network box finite element model based on the ring network box design parameter set; a multiple operating condition characteristic matrix determination module, used to perform scene parameter feature mining under multiple operating conditions based on the target deployment location and determine multiple operating condition characteristic matrices; and a multi-operating condition electric field anomaly map acquisition module, used to perform virtual mapping of the electric field under the ring network box finite element model based on the multiple operating condition characteristic matrices. The system includes: a distribution anomaly solution module to obtain multi-condition electric field anomaly maps; a ring network box design optimization guidance network establishment module to perform multi-parameter sensitivity coupling correlation calculations on the ring network box design parameter group based on the multi-condition electric field anomaly maps, and establish a ring network box design optimization guidance network; a ring network box design adjustment space acquisition module to perform feedback adjustment on the ring network box design parameter group based on the ring network box design optimization guidance network, and acquire the ring network box design adjustment space; and a ring network box design optimization result acquisition module to perform multi-condition collaborative electric field distribution risk optimization on the ring network box design adjustment space based on the multiple condition characteristic matrices, and acquire the ring network box design optimization result.

[0017] The technical solution provided in this application has at least the following technical effects or advantages: by achieving the technical goal of unified modeling and collaborative optimization analysis based on the coupling relationship between electric field characteristics and design parameters under multiple operating conditions, it achieves the technical effects of improving the accuracy of electric field anomaly identification and the scientific nature of design parameter optimization, enhancing the reliability of electrical insulation and the stability of operational safety.

[0018] The above description is merely an overview of the technical solution of this application. To enable a clearer understanding of the technical means of this application and to facilitate its implementation according to the description, and to make the above and other objects, features, and advantages of this application more apparent, specific embodiments of this application are described below. It should be understood that the content described in this section is not intended to identify key or important features of the embodiments of this application, nor is it intended to limit the scope of this application. Other features of this application will become readily apparent through the following description. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely exemplary. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0020] Figure 1This is a flowchart illustrating the method for optimizing the electric field distribution of the primary and secondary integrated ring network box in this application.

[0021] Figure 2 This is a schematic diagram of the electric field distribution optimization device for the primary and secondary integrated ring network box in this application.

[0022] Figure labeling: Module 1 for establishing the finite element model of the ring network box; Module 2 for determining the characteristic matrix of multiple working conditions; Module 3 for obtaining the electric field anomaly spectrum of multiple working conditions; Module 4 for establishing the ring network box design optimization guidance network; Module 5 for obtaining the adjustment space of the ring network box design; Module 6 for obtaining the optimization results of the ring network box design. Detailed Implementation

[0023] This application provides a method and apparatus for optimizing the electric field distribution of a primary and secondary integrated ring main unit (RMU). It addresses the technical problem in existing technologies where the lack of a unified modeling and collaborative analysis mechanism for the coupling relationship between multi-condition electric field distribution characteristics and design parameters makes it difficult to accurately identify and quantify abnormal electric field regions in complex operating environments. This further affects the scientific validity of parameter adjustments and the reliability of overall electrical insulation performance during design optimization, thus restricting the safe and stable operation of the RMU under complex electromagnetic environments. The application achieves the technical goal of unified modeling and collaborative optimization analysis based on the coupling relationship between multi-condition electric field characteristics and design parameters, thereby improving the accuracy of electric field anomaly identification, the scientific validity of design parameter optimization, and enhancing the reliability of electrical insulation and operational safety and stability.

[0024] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. It should be understood that this application is not limited to the exemplary embodiments described herein. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application. It should also be noted that, for ease of description, only the parts related to this application are shown in the accompanying drawings, not all of them.

[0025] Example 1, please refer to the appendix. Figure 1 This application provides a method for optimizing the electric field distribution of a primary and secondary integrated ring network box, which is applied to an electric field distribution optimization device for a primary and secondary integrated ring network box. The method specifically includes the following steps:

[0026] Obtain the design parameter set of the ring network box at the target deployment location, and establish a finite element model of the ring network box based on the design parameter set.

[0027] Furthermore, this application also includes: the ring main unit design parameter group includes a ring main unit structural parameter group, a ring main unit material parameter group, and a ring main unit electrical parameter group.

[0028] Specifically, obtaining the ring main unit design parameter set for the target deployment location refers to collecting and organizing a multi-dimensional set of parameters to characterize the overall design characteristics of the ring main unit, considering the spatial environmental conditions, electrical operating conditions, and installation constraints of the predetermined installation area. The target deployment location defines the geographical location, power grid topology, and external electromagnetic environment characteristics of the actual application scenario of the ring main unit. The ring main unit design parameter set describes the basic design information of the ring main unit in terms of structural configuration, material properties, and electrical performance. Specifically, the ring main unit structural parameter set characterizes the geometric structure information such as the internal conductor layout, insulation spacing, cavity dimensions, and component installation positions. The ring main unit material parameter set characterizes the physical properties of the conductor materials, insulating media, and shell materials, such as conductivity, dielectric constant, withstand voltage, and thermal stability. The ring main unit electrical parameter set characterizes the electrical operating characteristics such as rated voltage level, operating current range, voltage fluctuation characteristics, and grounding method. Through the classification, description, and unified modeling of structural, material, and electrical parameters, a comprehensive characterization and multi-dimensional expression of the ring main unit's design characteristics are achieved.

[0029] Furthermore, establishing a finite element model of the ring network box based on the design parameter set refers to constructing a discretized simulation model for numerical calculation based on the geometric structure, material properties, and electrical boundary conditions defined by the design parameters. The finite element model divides the continuous physical field region into multiple finite elements and applies corresponding physical control equations and constraints to each element to achieve accurate simulation and calculation of the electric field distribution characteristics inside the ring network box.

[0030] Based on the target deployment location, scene parameter feature mining is performed under multiple working conditions to determine multiple working condition characteristic matrices.

[0031] Furthermore, this application also includes: retrieving the power frequency voltage condition log set, lightning impulse condition log set, and partial discharge condition log set of the target deployment location; performing multi-parameter coupling relationship mining on the power frequency voltage condition log set to obtain a power frequency voltage scenario parameter association set; performing key parameter filtering and aggregation on the power frequency voltage condition log set based on the power frequency voltage scenario parameter association set to obtain a power frequency voltage condition characteristic matrix; performing multi-parameter coupling relationship mining and key parameter filtering and aggregation on the lightning impulse condition log set to obtain a lightning impulse condition characteristic matrix; performing multi-parameter coupling relationship mining and key parameter filtering and aggregation on the partial discharge condition log set to obtain a partial discharge condition characteristic matrix; and combining the power frequency voltage condition characteristic matrix and the lightning impulse condition characteristic matrix to generate the multiple condition characteristic matrices.

[0032] Specifically, retrieving the power frequency voltage condition log set, lightning impulse condition log set, and partial discharge condition log set for the target deployment location refers to extracting a collection of data records reflecting the changes in electrical state under different operating conditions from historical operating databases or online monitoring systems related to the target deployment location. The power frequency voltage condition log set records operating information such as voltage amplitude, current response, and load fluctuations under rated frequency conditions. The lightning impulse condition log set records transient voltage and current impulse waveforms and propagation characteristics under lightning overvoltage. The partial discharge condition log set records the characteristics and evolution of partial discharge signals generated by the insulation system under defects or stress. By uniformly retrieving these multiple types of condition logs, comprehensive data acquisition of the target deployment location's multi-scenario operating status can be achieved.

[0033] Furthermore, multi-parameter coupling relationship mining is performed on the power frequency voltage operating condition log set to obtain a power frequency voltage scenario parameter association set. This refers to analyzing the correlation between variables such as voltage, current, load, temperature, and environmental factors based on multi-dimensional monitoring parameters under power frequency operation conditions, forming a power frequency voltage scenario parameter association set that can characterize the parameter interaction characteristics under power frequency voltage operation scenarios. The power frequency voltage scenario parameter association set is used to describe the influence paths and intensity relationships between different parameters. Specifically, when mining multi-parameter coupling relationships on the power frequency voltage operating condition log set, the log set is first normalized and preprocessed, and a parameter time series set is constructed. The Pearson correlation coefficient is used to characterize the linear correlation between different parameters, mutual information is used to characterize the nonlinear statistical dependence between different parameters, and graph neural networks are used to characterize the higher-order topological relationships and propagation influence relationships between multiple parameters. Furthermore, addressing the issue of different output types of Pearson correlation coefficient, mutual information, and graph neural network, the linear correlation, nonlinear dependency, and graph topological correlation results are first normalized to map different dimensions and value ranges to a preset numerical interval. Then, based on preset fusion weights, the normalized multi-type correlations are weighted and fused to generate a unified parameter correlation value. Finally, the unified parameter correlation values ​​are matrix-arranged according to parameter correspondences to generate a working condition characteristic matrix. The Pearson correlation coefficient is used to calculate the degree of linear correlation between any two parameters. The formula for calculating the Pearson correlation coefficient is as follows: ρ ij =Cov(X i ,X j ) / (σX i ·σX j ), where ρ ij Cov(X) represents the correlation coefficient between the i-th parameter and the j-th parameter. i ,X j ) represents the covariance, σX i and σX jThese represent the standard deviations of the corresponding parameters. For nonlinear associations, mutual information is used to calculate the association strength. The mutual information calculation formula is as follows: I(X;Y)=∑p(x,y)log(p(x,y) / (p(x)p(y))). Wherein, I(X;Y) represents the mutual information value between parameter X and parameter Y. Furthermore, a graph neural network model is used to learn the higher-order coupling relationships between multiple parameters. Each parameter is input as a node into the graph convolutional network, and the degree of coupling influence between parameters is obtained through node feature propagation.

[0034] Furthermore, based on the parameter association set of the power frequency voltage scenario, key parameters are filtered and aggregated from the power frequency voltage operating condition log set to obtain the power frequency voltage operating condition characteristic matrix. This involves evaluating and filtering the importance of multi-dimensional parameters in the original log data according to parameter association relationships, extracting core parameters that significantly affect the electric field distribution, and arranging and representing them in a structured and matrix-like manner according to preset dimensions. This constructs a parameter matrix that reflects the main characteristics under power frequency voltage conditions, where the power frequency voltage operating condition characteristic matrix serves as the input feature carrier for subsequent electric field simulation and analysis. Further, during the key parameter filtering process, a feature importance evaluation method based on a random forest model is used to calculate the contribution of each parameter to the electric field anomaly indicators. Parameter importance value F. k The calculation formula is as follows: F k = Among them, ΔG k,t Let represent the decrease in the Gini index of the k-th parameter in the t-th decision tree, and T represent the number of decision trees. Further, parameters with importance values ​​greater than a preset importance threshold are identified as key parameters and aggregated according to operating condition categories to form corresponding operating condition characteristic matrices.

[0035] Furthermore, multi-parameter coupling relationship mining and key parameter screening and aggregation are performed on the lightning impact condition log set to obtain the lightning impact condition characteristic matrix. This refers to identifying the coupling relationship between relevant parameters based on the voltage surge, current surge and rapid change of electromagnetic field characteristics during the lightning transient process, and screening out key parameters that have a significant impact on electric field distribution and insulation performance. Then, a structured data model that can characterize the lightning impact condition is constructed through matrix method. The lightning impact condition characteristic matrix is ​​used to describe the electric field action characteristics under high-frequency transient conditions.

[0036] Furthermore, multi-parameter coupling relationship mining and key parameter screening and aggregation are performed on the partial discharge condition log set to obtain the partial discharge condition characteristic matrix. This involves analyzing the coupling relationships between relevant parameters based on key indicators such as the amplitude, frequency, phase distribution, and repetition characteristics of the partial discharge signal, and screening out core parameters that have a significant impact on local electric field distortion and insulation degradation. The partial discharge condition characteristic matrix is ​​then formed through a matrix organization method to characterize the electric field response features of the insulation system under defect conditions. Simultaneously, multiple condition characteristic matrices are generated by combining the power frequency voltage condition characteristic matrix and the lightning impulse condition characteristic matrix. This involves unifying and standardizing the characteristic matrices formed under different conditions to construct a comprehensive matrix set containing features of multiple operating scenarios, thereby achieving a unified expression and collaborative modeling of electric field analysis under multiple conditions.

[0037] Based on the multiple operating condition characteristic matrices, the electric field distribution anomaly is solved under virtual mapping of the finite element model of the ring network box, and the multi-operating condition electric field anomaly spectrum is obtained.

[0038] Furthermore, this application also includes: mapping the multiple operating condition characteristic matrices to the ring network box finite element model respectively to obtain multiple virtual operating condition mapping models; performing adaptive mesh partitioning and boundary condition loading on the ring network box finite element model according to each virtual operating condition mapping model to obtain multiple operating condition electric field solution models; performing steady-state electric field identification, transient impact electric field identification, and local distortion electric field identification on each operating condition electric field solution model to obtain multiple operating condition electric field distribution results; extracting local maximum field strength parameters, electric field gradient abrupt change parameters, insulation interface field strength parameters, and field strength concentration region parameters according to each operating condition electric field distribution result to obtain multiple electric field anomaly feature sets; and performing anomaly region labeling and map encoding according to the multiple electric field anomaly feature sets to obtain the multi-operating condition electric field anomaly map.

[0039] Specifically, the power frequency voltage characteristic matrix, lightning impulse characteristic matrix, and partial discharge characteristic matrix are mapped to the ring network box finite element model to obtain multiple virtual operating condition mapping models. This means that the voltage amplitude, frequency characteristics, impulse waveform parameters, and discharge characteristic parameters contained in the operating condition characteristic matrix are used as input variables and loaded into the ring network box finite element model through parameter assignment and scene driving. This constructs simulation scenarios corresponding to different operating states on the basis of unified geometric structure and material properties. The virtual operating condition mapping model is used to characterize the electric field distribution calculation environment under specific operating condition parameter constraints, and realizes the correspondence between multiple operating conditions and simulation models.

[0040] Furthermore, based on each virtual operating condition mapping model, the ring network box finite element model is subjected to adaptive meshing and boundary condition loading to obtain multiple operating condition electric field solution models. This means that, based on the virtual operating condition mapping model, the model region is discretized differently according to the electric field change gradient and geometric complexity. Adaptive meshing improves the calculation accuracy of key areas. At the same time, combined with the voltage source, grounding conditions and external electromagnetic constraints defined by the operating condition characteristic matrix, corresponding boundary conditions are applied to the model to form an electric field analysis model that meets the requirements of numerical solution. The operating condition electric field solution model is used to support the subsequent electric field numerical calculation process.

[0041] Furthermore, steady-state electric field identification, transient impact electric field identification, and local distortion electric field identification are performed on the electric field solution model for each operating condition to obtain multiple operating condition electric field distribution results. This refers to classifying and analyzing the electric field state with different time scales and physical characteristics based on the calculation results of the solution model. Among them, steady-state electric field identification is used to analyze the stability of electric field distribution under power frequency operation conditions, transient impact electric field identification is used to analyze the transient electric field change process caused by lightning impact or switching operation, and local distortion electric field identification is used to identify the local electric field concentration phenomenon caused by structural discontinuity or dielectric inhomogeneity. Spatial electric field distribution data under different operating conditions are obtained through multi-type electric field identification.

[0042] Furthermore, based on the electric field distribution results for each operating condition, parameters such as the local maximum field strength, electric field gradient abrupt change, insulation interface field strength, and field strength concentration region parameters are extracted to obtain multiple electric field anomaly feature sets. This refers to extracting key indicators that can reflect the abnormal state of the electric field from the electric field distribution results. Among them, the local maximum field strength parameter is used to characterize the extreme location and amplitude characteristics of the electric field strength, the electric field gradient abrupt change parameter is used to describe the abnormal region of the electric field change rate, the insulation interface field strength parameter is used to characterize the electric field distribution characteristics at the interface of different media, and the field strength concentration region parameter is used to characterize the degree of electric field concentration and spatial range. Through the extraction and combination of parameters, a multidimensional electric field anomaly feature set is formed.

[0043] Furthermore, based on multiple electric field anomaly feature sets, anomaly regions are labeled and graphically encoded to obtain multi-condition electric field anomaly maps. This involves identifying and classifying corresponding regions based on the spatial distribution and intensity characteristics of various anomaly indicators in the electric field anomaly feature sets, and then using graph structures or encoding methods to structurally express the anomaly regions and their relationships. This constructs a graphical model that can uniformly describe the distribution characteristics of electric field anomalies under different operating conditions. The multi-condition electric field anomaly map is used to achieve the visualization of anomaly regions and subsequent analysis and processing. Further, the graphical encoding adopts a two-layer node-edge structure. Nodes represent anomaly regions, electric field indicators, and design parameters, while edges represent regional relationships, parameter influence relationships, and operating condition relationships. The node encoding format is defined as: Node={node number, node type, spatial coordinates, anomaly level, indicator vector}; the edge encoding format is defined as: Edge={start node, end node, association weight, association type}. Further, a graph database is used to store the anomaly region map, and an adjacency index structure is used to achieve anomaly region association retrieval.

[0044] Based on the multi-condition electric field anomaly spectrum, multi-parameter sensitivity coupling correlation calculations are performed on the design parameter group of the ring network box to establish a ring network box design optimization guidance diagram.

[0045] Furthermore, this application also includes: extracting an electric field anomaly region set based on the multi-condition electric field anomaly map; determining multiple design parameter anomaly association pairs based on the positional correspondence and structural mapping relationship between the electric field anomaly region set and the ring network box design parameter group; calculating the sensitivity value of each design parameter in the ring network box design parameter group to local maximum field strength, electric field uniformity, insulation interface field strength, and partial discharge risk based on the multiple design parameter anomaly association pairs, and obtaining multiple design parameter sensitivity vectors; performing parameter coupling correlation analysis and cross-condition response consistency analysis on the ring network box design parameter group based on the multiple design parameter sensitivity vectors, and obtaining a design parameter coupling correlation matrix; and performing graph structure encoding based on the multiple design parameter sensitivity vectors and the design parameter coupling correlation matrix to generate the ring network box design optimization guidance graph.

[0046] Furthermore, this application also includes: the set of abnormal electric field regions includes a local maximum field strength abnormal region, an electric field gradient abrupt change region, an insulation interface abnormal region, and a partial discharge risk region.

[0047] Furthermore, this application also includes: extracting the Kth design parameter from the ring network box design parameter group, where K is a positive integer; extracting the target abnormal region and target abnormal index corresponding to the Kth design parameter based on the multiple design parameter anomaly correlation pairs; subjecting the Kth design parameter to a preset amplitude perturbation while keeping other parameters in the ring network box design parameter group unchanged, to obtain multiple ring network box design perturbation groups; mapping the multiple ring network box design perturbation groups to the ring network box finite element model respectively, and solving for the electric field distribution in the target abnormal region to obtain multiple perturbation electric field results; analyzing the multiple perturbation electric field results to obtain the electric field change sequence corresponding to the target abnormal index, wherein the electric field change sequence includes the local maximum field strength change, electric field uniformity change, insulation interface field strength change, and partial discharge risk change; calculating the multi-index sensitivity value corresponding to the Kth design parameter based on the electric field change sequence, and normalizing and vectorizing the multi-index sensitivity value to obtain the Kth design parameter sensitivity vector.

[0048] Specifically, extracting anomaly regions from multi-condition electric field anomaly maps involves filtering and grouping different types of anomaly regions from a map model containing electric field distribution information for multiple operating conditions, based on anomaly labeling results. The multi-condition electric field anomaly map describes the spatial distribution characteristics of electric field anomalies under various operating conditions such as power frequency, electromagnetic shock, and partial discharge. The anomaly region set summarizes spatial regions exhibiting characteristics of field strength anomalies, gradient anomalies, or insulation risks. This concentrated extraction of anomaly regions provides spatial location data for subsequent parameter correlation analysis. The anomaly region set includes locally maximum field strength anomaly regions, which refer to the spatial range where the electric field strength reaches a local maximum and is significantly higher than the surrounding area in the electric field distribution calculation results. These locally maximum field strength anomaly regions characterize the locations of electric field stress concentration and potential breakdown risks, typically related to conductor tip effects, structural discontinuities, or insufficient insulation distance. Identifying and extracting these regions can be used to assess the insulation safety margin of electrical equipment. The set of abnormal electric field regions includes regions of abrupt changes in electric field gradient, which are collections of areas where the electric field intensity changes rapidly in space and the rate of gradient change is significantly higher than the average level. These regions describe the discontinuous characteristics and abrupt changes in the electric field distribution, and typically appear at material interfaces, geometric abrupt changes, or electrode edges. Such regions are prone to causing localized electric stress concentration and electric field distortion, adversely affecting insulation performance. The set of abnormal electric field regions also includes abnormal regions at insulation interfaces, which are collections of areas where the electric field distribution is uneven due to differences in dielectric constant, conductivity, or interface defects at the junctions of different dielectric materials. These abnormal regions reflect the distribution characteristics of the electric field in multi-dielectric coupling environments and involve interfaces between solid and gas insulation or between different solid materials. These regions are prone to forming electric field concentrations or partial discharge initiation points, significantly impacting the reliability of the insulation system. The set of abnormal electric field regions includes partial discharge risk regions, which are the collection of potential regions that meet the conditions for partial discharge under the combined effects of electric field strength, spatial structure and material state. Partial discharge risk regions are used to characterize the locations in the insulation system where weak discharge activities may occur. These regions are usually closely related to the existence of air gaps, insulation defects or electric field distortion. By identifying partial discharge risk regions, it is possible to predict and warn of insulation degradation trends.

[0049] Furthermore, based on the positional correspondence and structural mapping relationship between the set of electric field anomaly regions and the design parameter group of the ring network box, multiple design parameter anomaly association pairs are determined. This means that based on the internal structural layout and parameter definition information of the ring network box, the position of the anomaly region in geometric space is matched with the corresponding structural unit and design parameter. The positional correspondence is used to describe the spatial mapping relationship between the anomaly region and the specific structural part, and the structural mapping relationship is used to describe the position and influence range of the design parameter in the physical structure. By establishing the association pair between the design parameter and the electric field anomaly through the dual mapping relationship, multiple design parameter anomaly association pairs are formed to characterize the parameter influence path.

[0050] Furthermore, the Kth design parameter is extracted from the ring main unit design parameter group. K is a positive integer, which refers to selecting one specific parameter as the analysis object from the parameter set containing multiple structural parameters, material parameters and electrical parameters according to the preset index order or numbering rules. The Kth design parameter is used to represent the single variable that is to be evaluated to affect the electric field distribution. K is a positive integer used to identify the parameter's position in the parameter group. By extracting the single design parameter one by one, the degree of influence of each parameter can be analyzed independently.

[0051] Furthermore, based on multiple design parameter anomaly correlation pairs, the target anomaly region and target anomaly index corresponding to the Kth design parameter are extracted. This means that among the established correlations between design parameters and electric field anomalies, anomaly regions and corresponding evaluation indicators that have a mapping relationship with the Kth design parameter are selected. The target anomaly region is used to limit the spatial range of the main influence of the Kth design parameter, and the target anomaly index is used as a quantitative indicator to characterize the degree of electric field anomaly, including local maximum field strength, electric field uniformity, insulation interface field strength, and partial discharge risk. The precise positioning of the analysis range is achieved by screening the correlations.

[0052] Furthermore, while keeping other parameters in the ring network box design parameter group unchanged, the Kth design parameter is subjected to a preset amplitude disturbance to obtain multiple ring network box design disturbance groups. This means that, under the premise of fixing the values ​​of all parameters except the Kth design parameter, multiple increases and decreases of different amplitudes are applied to the Kth design parameter. The other parameters are used as control variables to keep the state constant, and the preset amplitude disturbance is used to construct different parameter change scenarios. By changing the value of the Kth design parameter, a set of discrete design schemes is formed, thereby constructing multiple ring network box design disturbance groups for sensitivity analysis.

[0053] Furthermore, multiple ring network box design disturbance groups are mapped to the ring network box finite element model respectively, and the electric field distribution is solved for the target abnormal area to obtain multiple disturbance electric field results. This means loading different disturbance parameter combinations into the finite element model for numerical simulation and extracting the corresponding electric field distribution data in the target abnormal area. The disturbance electric field results are used to reflect the changes in electric field response under different parameter values. The electric field response data is collected by solving multiple disturbance scenarios.

[0054] Furthermore, multiple disturbance electric field results are analyzed to obtain the electric field change sequence corresponding to the target anomaly index. The electric field change sequence includes the local maximum field strength change, electric field uniformity change, insulation interface field strength change, and partial discharge risk change. This refers to performing differential or statistical analysis on the electric field results under each disturbance condition to calculate the index change amplitude relative to the baseline design state. The local maximum field strength change is used to represent the increase or decrease of the maximum electric field strength, the electric field uniformity change is used to represent the change in the dispersion of the electric field distribution, the insulation interface field strength change is used to represent the change in electric field stress at the interface, and the partial discharge risk change is used to represent the change in the probability or risk level of discharge. The electric field change sequence is formed by arranging the changes of each index in order of disturbance amplitude.

[0055] Furthermore, based on the electric field change sequence, the multi-index sensitivity value corresponding to the Kth design parameter is calculated, and the multi-index sensitivity value is normalized and vectorized to obtain the Kth design parameter sensitivity vector. This refers to calculating the response intensity of each index to parameter changes based on the functional relationship between the parameter disturbance amplitude and the corresponding electric field change. The calculation methods for the multi-index sensitivity value include, but are not limited to, using the difference ratio method or regression slope method for quantification. Specifically, for any index, the ratio of the disturbance amplitude change to the corresponding electric field change is selected as the sensitivity approximation value, or the slope of the fitting function is extracted as the sensitivity characterization value by performing linear or nonlinear fitting on the disturbance amplitude and the electric field change. The sensitivity values ​​of different indices are further normalized to eliminate the influence of dimensions and unify the scale range. Then, they are vectorized and arranged according to the index order to form a multi-dimensional vector expression including local maximum field strength sensitivity, electric field uniformity sensitivity, insulation interface field strength sensitivity, and partial discharge risk sensitivity, i.e., the Kth design parameter sensitivity vector. Furthermore, the sensitivity values ​​for multiple indicators are calculated using the local disturbance response function, for the Kth design parameter P. k With the target anomaly indicator E m Sensitivity S between k , m The following formula is used to calculate S: k , m =( E m / P k )·(P k / E m ), where S k , m This represents the normalized sensitivity value of the Kth design parameter to the mth anomaly index. For nonlinear coupling, a polynomial regression model is used to construct the nonlinear mapping relationship between parameter changes and electric field response: E m =a0+ Where ai represents the coefficient of the linear term, b ij The coefficients of the parameter coupling terms are used to determine the degree of coupling sensitivity between different parameters through the rate of change of the regression coefficients. Furthermore, to avoid fluctuations in results caused by a single difference perturbation, a multi-perturbation step-size averaging method is used to calculate the sensitivity value, and a weighted average is applied to the sensitivity results under different operating conditions, thereby improving the stability of the sensitivity analysis.

[0056] Furthermore, based on the sensitivity vectors of multiple design parameters, the design parameter group of the ring network box is subjected to parameter coupling correlation analysis and cross-condition response consistency analysis to obtain the design parameter coupling correlation matrix. This refers to the quantitative analysis of the mutual influence relationship between parameters based on the sensitivity performance of different design parameters under multiple indicators, and the evaluation of the stability influence capability of parameters in multi-condition environment by combining the consistency of parameter response change trends under different conditions. By matrix-expressing the correlation and synergistic change relationship between parameters, a design parameter coupling correlation matrix that can reflect the parameter coupling characteristics is formed.

[0057] Furthermore, graph structure encoding is performed based on multiple design parameter sensitivity vectors and design parameter coupling correlation matrices to generate a ring network box design optimization guidance graph. This refers to using design parameters as nodes, multiple design parameter sensitivity vectors to represent node attributes, and parameter coupling correlations as edge connections. A network structure reflecting the parameter influence paths and coupling relationships is constructed using graph structure modeling methods. Graph structure encoding is used to realize the structured expression of the relationships between parameters, and the ring network box design optimization guidance graph is used to guide the direction of parameter adjustment and the selection of optimization paths, thereby supporting the multi-parameter collaborative optimization process. Further, in the graph structure encoding process, the ring network box design parameters are defined as graph nodes, and the coupling correlations between parameters are defined as edge connections. The edge weights are determined by the correlation values ​​in the design parameter coupling correlation matrix. Any two parameter nodes V... i With V j The edge weight W between ij The calculation formula is as follows: W ij =αρ ij +βI ij +γC ij Wherein, ρ ij I represents the linear correlation coefficient.ij Represents the mutual information value, C ij The coefficient represents the consistency coefficient of response across operating conditions, and α, β, and γ represent the weighting coefficients. Further, the adjacency matrix A = [W...] is used. ij The graph structure is encoded using the node feature matrix H=[S i ] represents multiple design parameter sensitivity vectors corresponding to each design parameter, thereby constructing a design optimization guidance diagram for the ring network box.

[0058] The design parameters of the ring network box are adjusted based on the design optimization guidance diagram to obtain the design adjustment space of the ring network box.

[0059] Furthermore, this application also includes: fitting the electric field distribution of each ring network box design candidate scheme in the ring network box design candidate space according to the power frequency voltage condition characteristic matrix to obtain multiple power frequency voltage electric field distribution characteristics; training an electric field distribution risk prediction model based on the ring network box electric field distribution risk event set; inputting the multiple power frequency voltage electric field distribution characteristics into the electric field distribution risk prediction model to obtain multiple electric field distribution risk values; and filtering the ring network box design candidate space according to the electric field distribution risk threshold based on the multiple electric field distribution risk values ​​to obtain the first optimal ring network box design domain.

[0060] Specifically, the feedback adjustment of the ring network box design parameter group based on the ring network box design optimization guidance graph refers to using a graph structure model constructed with design parameters as nodes, coupling relationships between parameters as edges, and sensitivity information integrated, to analyze the influence path and intensity of each design parameter in the electric field distribution. The ring network box design optimization guidance graph is used to characterize the correlation between different parameters and their contribution to electric field anomaly indicators. By analyzing the node weights, edge connection strengths, and propagation relationships in the graph structure, the key parameters that need to be adjusted first and the adjustment direction are determined. Based on the sensitivity information of multiple indicators, the parameter values ​​are corrected and optimized, thereby achieving systematic feedback adjustment of the design parameter group.

[0061] Furthermore, obtaining the design adjustment space of the ring main unit refers to determining the feasible variation range and combination range of each design parameter after completing the feedback adjustment of the design parameters, combined with the design constraints, parameter value ranges, and coupling relationships between parameters. The design adjustment space of the ring main unit is used to describe the multi-dimensional parameter domain in which the design parameters can be adjusted under the premise of satisfying structural constraints, electrical performance constraints, and safety constraints. By jointly expressing the adjustment range of each parameter, a design space model that can be used for subsequent optimization calculations is formed.

[0062] Based on the multiple operating condition characteristic matrices, the design adjustment space of the ring network box is optimized by multi-operating condition collaborative electric field distribution risk search, and the ring network box design optimization results are obtained.

[0063] Furthermore, this application also includes: filtering the ring network box design adjustment space according to the ring network box design constraints at the target deployment location to obtain a ring network box design candidate space; optimizing the electric field distribution risk of the ring network box design candidate space according to the power frequency voltage condition characteristic matrix to obtain a first optimal ring network box design domain; optimizing the electric field distribution risk of the first optimal ring network box design domain according to the lightning impulse condition characteristic matrix to obtain a second optimal ring network box design domain; optimizing the electric field distribution risk of the second optimal ring network box design domain according to the partial discharge condition characteristic matrix to obtain a third optimal ring network box design domain; and optimizing the third optimal ring network box design domain by minimizing the electric field non-uniformity risk to generate the ring network box design optimization result.

[0064] Furthermore, this application also includes: fitting the electric field distribution of each ring network box design candidate scheme in the ring network box design candidate space according to the power frequency voltage condition characteristic matrix to obtain multiple power frequency voltage electric field distribution characteristics; training an electric field distribution risk prediction model based on the ring network box electric field distribution risk event set; inputting the multiple power frequency voltage electric field distribution characteristics into the electric field distribution risk prediction model to obtain multiple electric field distribution risk values; and filtering the ring network box design candidate space according to the electric field distribution risk threshold based on the multiple electric field distribution risk values ​​to obtain the first optimal ring network box design domain.

[0065] Specifically, filtering the ring main unit design adjustment space based on the design constraints of the target deployment location to obtain the ring main unit design candidate space means, on the basis of the already constructed multi-dimensional parameter adjustment range, introducing constraints related to the target deployment location, such as structural size restrictions, installation space restrictions, electrical insulation level requirements, and operational safety specifications, and screening and compliance judgments for each design parameter combination item by item. Among them, the ring main unit design constraints are used to limit the feasibility boundary of the parameter combination, and the ring main unit design adjustment space is used to describe the continuous variation range of parameters. By eliminating parameter combinations that do not meet the constraints, the set of parameter combinations that meet the design specifications is retained, thereby forming a ring main unit design candidate space consisting of multiple feasible design schemes.

[0066] Furthermore, based on the power frequency voltage condition characteristic matrix, the electric field distribution of each ring main unit design candidate scheme in the ring main unit design candidate space is fitted to obtain multiple power frequency voltage electric field distribution characteristics. This means using the voltage amplitude, frequency characteristics, and load-related parameters contained in the power frequency voltage condition characteristic matrix as input constraints, performing electromagnetic field numerical simulation or function fitting calculations on each design scheme in the ring main unit design candidate space, and reconstructing the electric field spatial distribution results under power frequency conditions through the finite element method or equivalent analytical model. The ring main unit design candidate space is used to characterize a set of multiple design schemes that meet the basic constraints, and the power frequency voltage electric field distribution characteristics are used to describe the electric field intensity distribution, gradient change, and spatial distribution law of each design scheme under steady-state power frequency operation conditions, thereby forming a set of basic characteristic data for subsequent risk assessment.

[0067] Furthermore, training an electric field distribution risk prediction model based on the electric field distribution risk event set of the ring network box refers to constructing a sample dataset based on electric field anomaly events and equipment failure events recorded in historical operating data. The electric field distribution risk event set of the ring network box is used to characterize risk events such as electric field anomalies, insulation breakdown, or partial discharge that occur under different design conditions and operating conditions. By performing feature extraction, labeling, and data preprocessing on the sample data, machine learning algorithms or statistical learning methods are used to model and train the mapping relationship between input features and risk results, thereby obtaining an electric field distribution risk prediction model that can assess the risks of unknown design schemes. Furthermore, due to the limited number of actual failure samples, a training sample set is constructed by combining historical operation monitoring data, on-site partial discharge detection data, insulation fault record data, and finite element simulation data. The historical operation monitoring data is used to reflect the electric field response characteristics under actual operating conditions, the on-site partial discharge detection data and insulation fault record data are used to provide labels for actual risk events, and the finite element simulation data is used to supplement the electric field distribution characteristics under extreme operating conditions and scenarios with few samples. By jointly training an electric field distribution risk prediction model with multi-source heterogeneous data, a mapping relationship between electric field characteristics and actual operating risks is established.

[0068] Furthermore, inputting multiple power frequency voltage electric field distribution characteristics into the electric field distribution risk prediction model to obtain multiple electric field distribution risk values ​​means inputting the electric field distribution characteristic data corresponding to each design candidate scheme under power frequency conditions as input variables into the trained electric field distribution risk prediction model. The electric field distribution risk prediction model comprehensively calculates the electric field intensity distribution characteristics, gradient change characteristics, and spatial uniformity characteristics, and outputs the electric field distribution risk quantification value of the corresponding design scheme. The electric field distribution risk value is used to characterize the probability of electric field anomalies or insulation failures occurring under power frequency operating conditions.

[0069] Furthermore, based on multiple electric field distribution risk values, the candidate space for ring main unit design is screened according to the electric field distribution risk threshold to obtain the first optimal ring main unit design domain. This means comparing and analyzing the electric field distribution risk values ​​corresponding to all design candidate schemes, retaining design schemes with electric field distribution risk values ​​less than the preset electric field distribution risk threshold, and eliminating design schemes with electric field distribution risk values ​​greater than or equal to the preset threshold, thereby forming a subset of design schemes that meet the power frequency safety constraints. The first optimal ring main unit design domain is used to represent the set of preferred design schemes whose risk level is within an acceptable range under the power frequency voltage condition constraints.

[0070] Furthermore, based on the lightning impact condition characteristic matrix, the electric field distribution risk of the first optimal ring main unit design domain is optimized to obtain the second optimal ring main unit design domain. This means that, based on the set of design schemes that already meet the power frequency operating conditions, transient characteristic parameters under lightning impact conditions are introduced, and the impact electric field response of each design scheme is analyzed. By constructing an electric field distribution risk prediction model for lightning impact scenarios, the peak electric field, shock wave propagation characteristics, and insulation stress level are evaluated. Based on the calculated electric field distribution risk value, the design schemes are screened again, eliminating schemes with high breakdown risk or electric field concentration risk under lightning impact conditions, and retaining the set of design schemes that meet the impact tolerance requirements, thus forming the second optimal ring main unit design domain.

[0071] Furthermore, based on the partial discharge condition characteristic matrix, the electric field distribution risk of the second optimal ring network box design domain is optimized to obtain the third optimal ring network box design domain. This means that, based on the set of design schemes that meet the power frequency and lightning impact conditions, the partial discharge behavior under insulation defects and electric field distortion conditions is further considered. By introducing the partial discharge condition characteristic matrix, the micro electric field distribution, interface electric field stress and discharge initiation conditions of each design scheme are analyzed. The probability of discharge occurrence or risk level is calculated by combining the partial discharge risk prediction model. The design schemes are screened according to the preset discharge risk threshold, and the set of design schemes with lower partial discharge risk is retained, thus forming the third optimal ring network box design domain.

[0072] Furthermore, the optimization of the third optimization ring network box design domain by minimizing the risk of electric field non-uniformity generates the ring network box design optimization result. This means that in the set of design schemes after multi-condition step-by-step screening, electric field uniformity is used as the optimization objective. By constructing an evaluation index function for electric field non-uniformity, the electric field distribution of each design scheme is quantitatively evaluated. An optimization algorithm is then used to sort or further search the design schemes, and the design scheme that makes the electric field distribution most uniform and the overall risk the least is selected as the final optimization result. Here, the risk of electric field non-uniformity is used to reflect the degree to which the electric field distribution deviates from the uniform state, and the ring network box design optimization result is used to guide the actual engineering design and parameter configuration.

[0073] In summary, the electric field distribution optimization method for primary and secondary integrated ring network boxes provided in this application has the following technical effects: by achieving the technical goal of unified modeling and collaborative optimization analysis based on the coupling relationship between electric field characteristics and design parameters under multiple operating conditions, it can improve the accuracy of electric field anomaly identification, the scientific nature of design parameter optimization, and enhance the reliability of electrical insulation and the stability of operational safety.

[0074] Example 2: Based on the same inventive concept as the primary and secondary fusion ring network box electric field distribution optimization method in the aforementioned examples, this application also provides a primary and secondary fusion ring network box electric field distribution optimization device. Please refer to the appendix. Figure 2 The system includes: a ring network box finite element model establishment module 1, used to obtain the ring network box design parameter set at the target deployment location and establish a ring network box finite element model based on the ring network box design parameter set; a multiple operating condition characteristic matrix determination module 2, used to perform scene parameter feature mining under multiple operating conditions based on the target deployment location and determine multiple operating condition characteristic matrices; a multi-operating condition electric field anomaly map acquisition module 3, used to solve the electric field distribution anomaly under virtual mapping of the ring network box finite element model based on the multiple operating condition characteristic matrices and obtain a multi-operating condition electric field anomaly map; and a ring network box design optimization guidance module. The network establishment module 4 is used to perform multi-parameter sensitivity coupling correlation calculations on the ring network box design parameter group based on the multi-condition electric field anomaly spectrum, and establish a ring network box design optimization guidance network; the ring network box design adjustment space acquisition module 5 is used to perform feedback adjustment on the ring network box design parameter group based on the ring network box design optimization guidance network, and acquire the ring network box design adjustment space; the ring network box design optimization result acquisition module 6 is used to perform multi-condition collaborative electric field distribution risk optimization on the ring network box design adjustment space based on the multiple condition characteristic matrices, and acquire the ring network box design optimization result.

[0075] Furthermore, the primary and secondary fusion ring network box electric field distribution optimization device is also used for: retrieving the power frequency voltage condition log set, lightning impulse condition log set, and partial discharge condition log set at the target deployment location; performing multi-parameter coupling relationship mining on the power frequency voltage condition log set to obtain a power frequency voltage scenario parameter association set; performing key parameter filtering and aggregation on the power frequency voltage condition log set according to the power frequency voltage scenario parameter association set to obtain a power frequency voltage condition characteristic matrix; performing multi-parameter coupling relationship mining and key parameter filtering and aggregation on the lightning impulse condition log set to obtain a lightning impulse condition characteristic matrix; performing multi-parameter coupling relationship mining and key parameter filtering and aggregation on the partial discharge condition log set to obtain a partial discharge condition characteristic matrix; and combining the power frequency voltage condition characteristic matrix and the lightning impulse condition characteristic matrix to generate the multiple condition characteristic matrices.

[0076] Furthermore, the primary and secondary fusion ring network box electric field distribution optimization device is also used for: mapping the multiple operating condition characteristic matrices to the ring network box finite element model respectively to obtain multiple virtual operating condition mapping models; performing mesh adaptive partitioning and boundary condition loading on the ring network box finite element model according to each virtual operating condition mapping model to obtain multiple operating condition electric field solution models; performing steady-state electric field identification, transient impact electric field identification, and local distortion electric field identification on each operating condition electric field solution model to obtain multiple operating condition electric field distribution results; extracting local maximum field strength parameters, electric field gradient abrupt change parameters, insulation interface field strength parameters, and field strength concentration region parameters according to each operating condition electric field distribution result to obtain multiple electric field anomaly feature sets; and performing anomaly region labeling and map encoding according to the multiple electric field anomaly feature sets to obtain the multi-operating condition electric field anomaly map.

[0077] Furthermore, the primary and secondary fusion ring network box electric field distribution optimization device is also used for: extracting an electric field anomaly region set based on the multi-condition electric field anomaly map; determining multiple design parameter anomaly association pairs based on the positional correspondence and structural mapping relationship between the electric field anomaly region set and the ring network box design parameter group; calculating the sensitivity value of each design parameter in the ring network box design parameter group to local maximum field strength, electric field uniformity, insulation interface field strength, and partial discharge risk based on the multiple design parameter anomaly association pairs, and obtaining multiple design parameter sensitivity vectors; performing parameter coupling correlation analysis and cross-condition response consistency analysis on the ring network box design parameter group based on the multiple design parameter sensitivity vectors, and obtaining a design parameter coupling correlation matrix; and performing graph structure encoding based on the multiple design parameter sensitivity vectors and the design parameter coupling correlation matrix to generate the ring network box design optimization guidance graph.

[0078] Furthermore, the primary and secondary fusion ring network box electric field distribution optimization device is also used for: extracting the Kth design parameter from the ring network box design parameter group, where K is a positive integer; extracting the target abnormal region and target abnormal index corresponding to the Kth design parameter based on the multiple design parameter anomaly correlation pairs; subjecting the Kth design parameter to a preset amplitude perturbation while keeping other parameters in the ring network box design parameter group unchanged, to obtain multiple ring network box design perturbation groups; mapping the multiple ring network box design perturbation groups to the ring network box finite element model respectively, and solving the electric field distribution for the target abnormal region to obtain multiple perturbation electric field results; analyzing the multiple perturbation electric field results to obtain the electric field change sequence corresponding to the target abnormal index, the electric field change sequence including the local maximum field strength change, electric field uniformity change, insulation interface field strength change, and partial discharge risk change; calculating the multi-index sensitivity value corresponding to the Kth design parameter based on the electric field change sequence, and normalizing and vectorizing the multi-index sensitivity value to obtain the Kth design parameter sensitivity vector.

[0079] Furthermore, the primary and secondary fusion ring main unit electric field distribution optimization device is also used for: filtering the ring main unit design adjustment space according to the ring main unit design constraints at the target deployment location to obtain a ring main unit design candidate space; performing electric field distribution risk optimization on the ring main unit design candidate space according to the power frequency voltage condition characteristic matrix to obtain a first optimal ring main unit design domain; performing electric field distribution risk optimization on the first optimal ring main unit design domain according to the lightning impulse condition characteristic matrix to obtain a second optimal ring main unit design domain; performing electric field distribution risk optimization on the second optimal ring main unit design domain according to the partial discharge condition characteristic matrix to obtain a third optimal ring main unit design domain; and performing minimum electric field non-uniformity risk optimization on the third optimal ring main unit design domain to generate the ring main unit design optimization result.

[0080] Furthermore, the primary and secondary fusion ring network box electric field distribution optimization device is also used for: fitting the electric field distribution of each ring network box design candidate scheme in the ring network box design candidate space according to the power frequency voltage condition characteristic matrix, and obtaining multiple power frequency voltage electric field distribution characteristics; training an electric field distribution risk prediction model according to the ring network box electric field distribution risk event set; inputting the multiple power frequency voltage electric field distribution characteristics into the electric field distribution risk prediction model to obtain multiple electric field distribution risk values; and filtering the ring network box design candidate space according to the electric field distribution risk threshold based on the multiple electric field distribution risk values ​​to obtain the first optimal ring network box design domain.

[0081] Furthermore, the primary and secondary integrated ring network box electric field distribution optimization device is also used for: the ring network box design parameter group includes the ring network box structural parameter group, the ring network box material parameter group, and the ring network box electrical parameter group.

[0082] Furthermore, the primary and secondary fusion ring network box electric field distribution optimization device is also used for: the electric field anomaly region set includes local maximum field strength anomaly region, electric field gradient abrupt change region, insulation interface anomaly region and partial discharge risk region.

[0083] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The electric field distribution optimization method and specific examples of the primary and secondary fusion ring network box in the aforementioned embodiment one are also applicable to the electric field distribution optimization device of the primary and secondary fusion ring network box in this embodiment. Through the foregoing detailed description of the electric field distribution optimization method of the primary and secondary fusion ring network box, those skilled in the art can clearly understand the electric field distribution optimization device of the primary and secondary fusion ring network box in this embodiment. Therefore, for the sake of brevity, it will not be described in detail here.

[0084] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0085] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of this application and its equivalents, this application also intends to include such modifications and variations.

Claims

1. A method for optimizing the electric field distribution of a primary and secondary integrated ring network box, characterized in that, The method includes: Obtain the design parameter set of the ring network box at the target deployment location, and establish a finite element model of the ring network box based on the design parameter set; Based on the target deployment location, scene parameter feature mining is performed under multiple working conditions to determine multiple working condition characteristic matrices; Based on the multiple operating condition characteristic matrices, the electric field distribution anomaly of the ring network box finite element model is solved under virtual mapping to obtain the multi-operating condition electric field anomaly spectrum. Based on the multi-condition electric field anomaly spectrum, multi-parameter sensitivity coupling correlation calculations are performed on the design parameter group of the ring network box to establish a ring network box design optimization guidance diagram. Based on the ring network box design optimization guidance diagram, the design parameter group of the ring network box is adjusted to obtain the ring network box design adjustment space; Based on the multiple operating condition characteristic matrices, the design adjustment space of the ring network box is optimized by multi-operating condition collaborative electric field distribution risk search, and the ring network box design optimization results are obtained.

2. The method for optimizing the electric field distribution of a primary and secondary integrated ring network box as described in claim 1, characterized in that, Based on the target deployment location, scene parameter feature mining is performed under multiple operating conditions to determine multiple operating condition characteristic matrices, including: Retrieve the power frequency voltage condition log set, lightning impulse condition log set, and partial discharge condition log set of the target deployment location; Multi-parameter coupling relationship mining is performed on the power frequency voltage condition log set to obtain the power frequency voltage scenario parameter association set; Based on the power frequency voltage scenario parameter association set, the power frequency voltage operating condition log set is filtered and aggregated for key parameters to obtain the power frequency voltage operating condition characteristic matrix. Multi-parameter coupling relationship mining and key parameter filtering and aggregation are performed on the lightning impact condition log set to obtain the lightning impact condition characteristic matrix; Multi-parameter coupling relationship mining and key parameter filtering and aggregation are performed on the partial discharge condition log set to obtain the partial discharge condition characteristic matrix. The multiple condition characteristic matrices are generated by combining the power frequency voltage condition characteristic matrix and the lightning impulse condition characteristic matrix.

3. The method for optimizing the electric field distribution of a primary and secondary integrated ring network box as described in claim 1, characterized in that, Based on the multiple operating condition characteristic matrices, the electric field distribution anomaly is solved under virtual mapping of the finite element model of the ring network box, and multi-condition electric field anomaly maps are obtained, including: The multiple operating condition characteristic matrices are mapped to the ring network box finite element model respectively to obtain multiple virtual operating condition mapping models; Based on each virtual working condition mapping model, the ring network box finite element model is subjected to adaptive meshing and boundary condition loading to obtain multiple working condition electric field solution models. For each operating condition electric field solution model, steady-state electric field identification, transient impact electric field identification, and local distortion electric field identification are performed to obtain multiple operating condition electric field distribution results; Based on the electric field distribution results for each operating condition, extract the local maximum field strength parameter, electric field gradient abrupt change parameter, insulation interface field strength parameter, and field strength concentration region parameter to obtain multiple electric field anomaly feature sets. Based on the multiple electric field anomaly feature sets, anomaly regions are labeled and mapped to obtain the multi-condition electric field anomaly map.

4. The method for optimizing the electric field distribution of a primary and secondary integrated ring network box as described in claim 1, characterized in that, Based on the multi-condition electric field anomaly spectrum, multi-parameter sensitivity coupling correlation calculations are performed on the ring network box design parameter group to establish a ring network box design optimization guidance network, including: Extract the set of electric field anomaly regions based on the multi-condition electric field anomaly map; Based on the positional correspondence and structural mapping relationship between the set of abnormal electric field regions and the design parameter group of the ring network box, multiple abnormal design parameter association pairs are determined; Based on the multiple abnormal correlation pairs of design parameters, the sensitivity value of each design parameter in the ring network box design parameter group to local maximum field strength, electric field uniformity, insulation interface field strength and partial discharge risk is calculated, and multiple design parameter sensitivity vectors are obtained. Based on the multiple design parameter sensitivity vectors, perform parameter coupling correlation analysis and cross-condition response consistency analysis on the ring network box design parameter group to obtain the design parameter coupling correlation matrix; The design optimization guidance graph of the ring network box is generated by encoding the graph structure based on the sensitivity vectors of the multiple design parameters and the coupling correlation matrix of the design parameters.

5. The method for optimizing the electric field distribution of a primary and secondary integrated ring network box as described in claim 4, characterized in that, Based on the aforementioned multiple design parameter anomaly correlation pairs, the sensitivity value of each design parameter in the ring network enclosure design parameter group to local maximum electric field strength, electric field uniformity, insulation interface electric field strength, and partial discharge risk is calculated, obtaining multiple design parameter sensitivity vectors, including: The Kth design parameter is extracted from the ring network box design parameter group, where K is a positive integer; Based on the multiple design parameter anomaly association pairs, extract the target anomaly region and target anomaly index corresponding to the Kth design parameter; While keeping other parameters in the ring network box design parameter group unchanged, the Kth design parameter is subjected to a preset amplitude disturbance to obtain multiple ring network box design disturbance groups; The multiple ring network box design disturbance groups are respectively mapped to the ring network box finite element model, and the electric field distribution is solved for the target abnormal region to obtain multiple disturbance electric field results; The results of the multiple disturbance electric fields are analyzed to obtain the electric field change sequence corresponding to the target anomaly index. The electric field change sequence includes the local maximum field strength change, the electric field uniformity change, the field strength change at the insulation interface, and the partial discharge risk change. Based on the electric field change sequence, calculate the multi-index sensitivity value corresponding to the Kth design parameter, and normalize and vectorize the multi-index sensitivity value to obtain the sensitivity vector of the Kth design parameter.

6. The method for optimizing the electric field distribution of a primary and secondary integrated ring network box as described in claim 1, characterized in that, Based on the multiple operating condition characteristic matrices, the design adjustment space of the ring network box is optimized by multi-operating condition collaborative electric field distribution risk assessment to obtain the ring network box design optimization results, including: Based on the design constraints of the ring network box at the target deployment location, filter the ring network box design adjustment space to obtain the ring network box design candidate space; Based on the power frequency voltage condition characteristic matrix, the electric field distribution risk optimization is performed on the candidate space of the ring network box design to obtain the first optimal ring network box design domain. Based on the lightning impact condition characteristic matrix, the electric field distribution risk optimization is performed on the first optimal ring network box design domain to obtain the second optimal ring network box design domain. Based on the partial discharge condition characteristic matrix, the electric field distribution risk optimization is performed on the second optimal ring network box design domain to obtain the third optimal ring network box design domain. Minimize the risk of electric field non-uniformity in the third optimization ring network box design domain to generate the optimization result of the ring network box design.

7. The method for optimizing the electric field distribution of a primary and secondary integrated ring network box as described in claim 6, characterized in that, Based on the power frequency voltage operating condition characteristic matrix, the electric field distribution risk optimization is performed on the candidate design space of the ring main unit to obtain the first optimal ring main unit design domain, including: Based on the power frequency voltage operating condition characteristic matrix, the electric field distribution of each ring network box design candidate scheme in the ring network box design candidate space is fitted to obtain multiple power frequency voltage electric field distribution characteristics. Based on the set of risk events related to the electric field distribution of the ring network box, a risk prediction model for the electric field distribution is trained. The electric field distribution characteristics of the multiple power frequency voltages are input into the electric field distribution risk prediction model to obtain multiple electric field distribution risk values. Based on the multiple electric field distribution risk values, the candidate space for ring network box design is screened according to the electric field distribution risk threshold to obtain the first optimal ring network box design domain.

8. The method for optimizing the electric field distribution of a primary and secondary integrated ring network box as described in claim 1, characterized in that, The ring main unit design parameter group includes the ring main unit structural parameter group, the ring main unit material parameter group, and the ring main unit electrical parameter group.

9. The method for optimizing the electric field distribution of a primary and secondary integrated ring network box as described in claim 4, characterized in that, The set of abnormal electric field regions includes regions with local maximum electric field strength anomalies, regions with abrupt changes in electric field gradients, regions with abnormal insulation interfaces, and regions with partial discharge risk.

10. A device for optimizing the electric field distribution of a primary and secondary integrated ring network box, characterized in that, The steps for implementing the primary and secondary fusion ring network box electric field distribution optimization method according to any one of claims 1 to 9 include: The ring network box finite element model establishment module is used to obtain the ring network box design parameter set at the target deployment location and establish the ring network box finite element model based on the ring network box design parameter set; A module for determining multiple working condition characteristic matrices is used to perform scene parameter feature mining under multiple working conditions based on the target deployment location, and to determine multiple working condition characteristic matrices. The multi-condition electric field anomaly map acquisition module is used to solve the electric field distribution anomaly under the virtual mapping of the ring network box finite element model based on the multiple condition characteristic matrices, and to obtain the multi-condition electric field anomaly map. The ring network box design optimization guidance network establishment module is used to perform multi-parameter sensitivity coupling correlation calculation on the ring network box design parameter group based on the multi-condition electric field anomaly spectrum, and establish the ring network box design optimization guidance network; The ring network box design adjustment space acquisition module is used to adjust the ring network box design parameter group based on the ring network box design optimization guidance diagram to acquire the ring network box design adjustment space. The ring network box design optimization result acquisition module is used to perform multi-condition collaborative electric field distribution risk optimization on the ring network box design adjustment space based on the multiple operating condition characteristic matrices, and obtain the ring network box design optimization result.

Images