Power distribution network feasibility research report automatic generation method based on LDA probability model

Through the automatic generation method of feasibility study report of the distribution network based on the LDA probability model, the problem of redundancy assessment and management of the distribution network is solved, the recovery ability and resource utilization of the power grid are improved, and the optimization of distribution network management is achieved.

CN119047432AActive Publication Date: 2024-11-29GUANGDONG SHUNDE POWER DESIGN INSTITUTE CO LTD

Patent Information

Application Number
CN202411115568.8
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-14
Publication Date
2024-11-29
Estimated Expiration
2044-08-14

Smart Images

  • Figure CN119047432A_ABST
    Figure CN119047432A_ABST
Patent Text Reader

Abstract

The invention provides a power distribution network feasibility research report automatic generation method based on an LDA probability model, and the method comprises the steps: obtaining the ledger information of power grid equipment and the electrical topology connection relation of the ledger information, and generating a ledger document containing the name, type, position and connection relation information of the equipment; constructing a graph structure of a reference power grid according to the obtained machine account attribute and the electrical topological relation, marking data of a standby line and standby equipment, and generating a marking document containing a reference power grid structure and marking information; an LDA model is trained, the number of topics is set, iterative training is carried out, and topic distribution of each document and vocabulary distribution of each topic are extracted; based on the redundancy information, a feasibility report generation model is constructed, and the feasibility report generation model is used for analyzing potential themes and modes in the redundancy information and generating a power distribution network redundancy feasibility report.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of information technology, and in particular to a method for automatically generating a feasibility study report of a distribution network based on an LDA probability model. Background Art

[0002] It is crucial to evaluate the redundancy of the distribution network because it is directly related to the reliability and resilience of the power system. It can quickly restore power supply through alternative paths when a fault occurs, reduce the impact of power outages, ensure uninterrupted social and economic activities, and enhance the overall power grid's ability to resist emergencies, ensure the safety and stability of power supply, and lay a solid foundation for sustainable economic development. However, the challenges faced by the redundancy evaluation of the distribution network include the large differences in distribution networks in different regions, different equipment models and parameters, which leads to many challenges in the construction of the benchmark power grid model. Furthermore, although high redundancy can improve the reliability of the distribution network, it also means a reduction in resource utilization. How to balance reliability and economy and find a better level of redundancy. Secondly, generating a feasibility report on the redundancy of the distribution network requires the help of advanced algorithm models and data mining technologies, such as the LDA probability model, and how to generate reports intelligently is also a technical difficulty. Therefore, how to systematically solve these problems, that is, how to effectively deal with the diversity of model construction, the depth requirements of data analysis, the integrated innovation of algorithm application, and the intelligent needs of report generation while taking into account the accuracy, efficiency and practicality of the evaluation, is a technical problem that needs to be solved urgently. Summary of the invention

[0003] The present invention provides a method for automatically generating a feasibility study report of a distribution network based on an LDA probability model, which mainly includes:

[0004] Obtain the ledger information of power grid equipment and its electrical topology connection relationship, and generate a ledger document containing equipment name, type, location, and connection relationship information;

[0005] The obtained ledger attributes and electrical topology relationships are used to construct a graph structure of the benchmark power grid, and the backup lines and backup equipment data are annotated to generate an annotation document containing the benchmark power grid structure and annotation information;

[0006] According to the backup line and backup equipment data in the benchmark power grid structure, combined with the number of alternative paths between different equipment and lines, the redundancy of each node in the distribution network is calculated, and an electrical topology diagram containing redundancy information and a redundancy information document containing redundancy calculation results is generated;

[0007] Based on the redundancy index calculated by the nodes, analyze the fault recovery capability of the distribution network at different redundancy levels, evaluate its switching and recovery speed when a fault occurs, and generate a fault analysis document containing the fault recovery capability analysis results and the switching and recovery speed evaluation results when a fault occurs;

[0008] Based on the historical fault data of the distribution network and the redundancy index, the resource utilization of the distribution network is evaluated, the idleness of the backup equipment and lines in the non-fault state is determined, and an evaluation document containing the historical fault data and resource utilization evaluation results is generated;

[0009] Summarize the ledger documents, annotation documents, redundancy information documents, fault analysis documents and evaluation documents of power grid equipment, perform preprocessing, build a vocabulary, represent each document as the word frequency vector of the word in the vocabulary, generate a document-word frequency matrix, the rows of the matrix represent documents, the columns represent words, and the values ​​in the matrix represent the frequency of the words in the document;

[0010] Train the LDA model, set the number of topics, perform iterative training, and extract the topic distribution of each document and the vocabulary distribution of each topic;

[0011] Based on the redundancy information, a feasibility report generation model is constructed. The feasibility report generation model is used to analyze the potential topics and patterns in the redundancy information and generate a distribution network redundancy feasibility report.

[0012] The technical solution provided by the embodiment of the present invention may have the following beneficial effects:

[0013] The present invention discloses a method for automatically generating a feasibility study report for a distribution network based on an LDA probability model. The method records the ledger information of power grid equipment and its electrical topological connection relationship in detail, and generates a ledger document containing the equipment name, type, location, and connection relationship, thereby solving the problem that the equipment information is dispersed and difficult to manage in the traditional method. A labeling document containing a reference power grid structure and labeling information is constructed, and then the redundancy of each node is calculated based on the backup line and equipment data in the labeling document. The redundancy information is integrated into the electrical topology diagram to form a detailed redundancy information document. The redundancy of each node in the distribution network can be calculated and the fault recovery capability of the distribution network can be evaluated accordingly. Furthermore, the present invention uses historical fault data and redundancy indicators to jointly evaluate the resource utilization of the distribution network and generate an evaluation document. It not only records historical fault data, but also analyzes the idle conditions of spare equipment and lines in non-fault states, providing a scientific basis for power grid management; based on the above documents, the present invention generates a document-word frequency matrix by constructing a vocabulary and representing the documents as word frequency vectors, and then trains the LDA model to extract the topic distribution and word distribution of the documents; the present invention generates a distribution network redundancy feasibility report, including redundancy calculation results, resource utilization analysis and feasibility evaluation, and performs rule verification. This report helps power grid managers adjust and maintain basic data and graph data, thereby improving the operating efficiency and stability of the power grid; in short, the present invention integrates and analyzes power grid data, which not only improves the management efficiency and transparency of the power grid, but also enhances the fault response capability and resource allocation efficiency of the power grid, thereby realizing the optimization and upgrading of distribution network management. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 The present invention is a flow chart of a method for automatically generating a feasibility study report for a distribution network based on an LDA probability model.

[0015] Figure 2 It is a schematic diagram of a method for automatically generating a feasibility study report of a distribution network based on an LDA probability model according to the present invention. DETAILED DESCRIPTION

[0016] The following will describe the technical solutions in the embodiments of the present invention in detail in conjunction with the accompanying drawings in the embodiments of the present invention. The described embodiments are only part of the embodiments of the present invention.

[0017] like Figure 1-2 In this embodiment, a method for automatically generating a feasibility study report of a distribution network based on an LDA probability model may specifically include:

[0018] S101. Obtain the inventory information of power grid equipment and its electrical topology connection relationship, and generate an inventory document containing the equipment name, type, location, and connection relationship information.

[0019] Obtain and analyze the drawing model file, extract the equipment symbols and text annotation information in the drawing model file, obtain the attribute information of the name, type, and location of the power grid equipment, and construct the equipment attribute table; analyze the graphic elements of the connecting lines and arrows in the drawing model file, combine the equipment location information in the equipment attribute table, determine the electrical connection relationship between the equipment, and construct the equipment topology connection relationship table; integrate the equipment attribute table and the equipment topology connection relationship table to form a ledger document; perform attribute analysis on the generated ledger document; based on the attribute analysis results, mine the implicit relationship and frequent patterns between the attributes, and find out the relevant attribute combinations that can form contextual logical connections and thinking chains; use the relevant attribute combinations found, combined with the business characteristics and professional knowledge of the power grid equipment, to expand and / or refine the ledger document; perform logic checks and consistency checks on the expanded and / or refined ledger documents.

[0020] Specifically, according to the model file exported by the Southern Power Grid Zhikan system, the model file is analyzed, and the name, type, location and other attribute information of the power grid equipment are obtained by extracting the equipment symbols, text annotations and other information in the model, and the equipment attribute table is constructed. Using the location information in the equipment attribute table, by analyzing the connecting lines, arrows and other graphic elements in the model file, the electrical connection relationship between the equipment is judged, and the equipment topology connection relationship table is constructed to determine the upstream and downstream relationship and electrical connectivity between the equipment. According to the equipment attribute table and the topology connection relationship table, data fusion technology, such as data mapping, association analysis and other methods, is used to integrate the attribute information and connection relationship, generate a complete structured data set of power grid equipment ledger information and electrical topology relationship, and form a ledger document. The attribute information such as equipment name, type, location, connection relationship, etc. is directly extracted from the structured ledger document without the use of natural language processing technology. Using the extracted attribute information, an association rule mining algorithm, such as the Apriori algorithm or the FP-growth algorithm, is used to discover the implicit relationship and frequent pattern between the attributes. By analyzing the association rules of attributes such as equipment type, location, and connection relationship, we can find out the relevant attribute combinations that can form contextual logical connections and thinking chains. Based on the discovered relevant attribute combinations, combined with the business characteristics and professional knowledge of power grid equipment, the ledger documents are expanded and refined. According to the association between equipment type and location, the operating status information of the equipment is supplemented; according to the association between equipment connection relationship and name, the business attribute information such as equipment maintenance record and fault history is supplemented to enrich the content of the ledger document. The knowledge graph is constructed for the expanded ledger document, and concepts and entities such as equipment, attributes, and relationships are defined to establish semantic links between them. Inference rules are formulated, such as constraint rules for equipment type and operating status, and inference rules for equipment connection relationship and fault propagation. Using knowledge graphs and inference rules, the expanded ledger documents are logically checked and consistent, and the logical relationship between attributes is verified to ensure the integrity and accuracy of the ledger documents, so as to form high-quality ledger documents of power grid equipment ledger information and electrical topology relationships. First, the OpenCV image processing library is used to preprocess the image model files exported by the Southern Power Grid Zhikan system, such as grayscale and binarization, to improve image quality. Then, the region-based convolutional neural network (R-CNN) is used to detect and identify the equipment symbols in the image file. Through the trained equipment symbol model, transformers, circuit breakers, busbars and other equipment are identified with an accuracy of more than 95%. Then, optical character recognition (OCR) technology, such as the Tesseract OCR engine, is used to extract the text annotation information in the image file, associate the equipment name, type and other attributes with the corresponding equipment symbols, and construct the equipment attribute table. For the identified connecting lines and arrows, the Hough transform algorithm is used for straight line detection. According to the starting and ending positions and directions of the connecting lines between the devices, the upstream and downstream topological relationship of the equipment is judged, and the equipment topological connection matrix is ​​generated.The equipment attribute table and the topological connection matrix are associated, and a structured data set containing detailed equipment attributes and connection relationships is generated through primary key matching, foreign key association, etc. The data set is mined for association rules. The Apriori algorithm is used to set the minimum support to 10% and the minimum confidence to 80%. The association rules between equipment type and voltage level are found, such as "transformer=>220kV", "circuit breaker=>110kV", etc., revealing the intrinsic connection between equipment attributes. According to the association rules, the ledger document is expanded, such as supplementing the capacity, ratio, wiring group and other parameters of the transformer, and supplementing the rated current, rated breaking current and other parameters of the circuit breaker. The Neo4j graph database is used to build the equipment knowledge graph, and the information such as equipment, attributes, connection relationships, etc. is stored in the form of nodes and edges. The query, reasoning and verification of equipment attributes and topological relationships are realized through the Cypher query language, such as "MATCH(d:Device)--.

[0021] >(t:Transformer)WHEREd.type='Circuit breaker'RETURNd,t" can query all circuit breakers directly connected to the transformer. Through the TransE algorithm, the entities and relationships in the knowledge graph are vectorized, and the representation learning model is used to predict and complete the equipment attributes and connection relationships, further improving the completeness and accuracy of the ledger documents.

[0022] S102, constructing a graph structure of a benchmark power grid with the acquired ledger attributes and electrical topology relationships, marking the backup lines and backup equipment data, and generating a marked document containing the benchmark power grid structure and marked information.

[0023] According to the extracted ledger attributes and electrical topological relationships, nodes and relationships are created, and the equipment is used as nodes and the topological connection relationship as an edge to construct a graph structure of the benchmark power grid, including marking the corresponding labels for the backup lines and backup equipment to distinguish between the main lines and backup lines and the main equipment and backup equipment, and marking the corresponding main line information for the backup lines to form a backup relationship mapping; analyzing the key nodes and key paths in the power grid topological structure, evaluating the importance and influence range of the nodes, and generating network characteristic parameters for the power supply path and load distribution according to the topological structure and equipment attributes; for the equipment nodes in the graph structure, the ledger attribute information is associated, and for the backup equipment, the attributes of its commissioning conditions and switching methods are associated to form a backup equipment attribute portrait; integrating the graph structure with the geographic information of the equipment, mapping the nodes to the map according to the geographic location attributes of the equipment, forming a power grid geographic topology map to display the spatial distribution and geographic wiring relationship of the power grid; based on the constructed graph structure of the benchmark power grid, generating a labeling document containing equipment nodes, topological connections, and backup labeling information, and the labeling document includes distinguishing different types of equipment and connections through the visual attributes of the node color and shape.

[0024] Specifically, according to the extracted ledger attributes and electrical topological relationships, the Neo4j graph database management tool is used to create nodes and relationships, with devices as nodes and topological connection relationships as edges to build the graph structure of the benchmark power grid. The Cypher query language is used to achieve flexible query and analysis of the power grid structure. In the process of building the graph structure, special label attributes such as "spare" and "redundant" are set for backup lines and backup equipment. Through attribute annotation, the main and backup equipment are distinguished to facilitate subsequent identification and management. For backup lines with corresponding main lines, the corresponding main line ID is marked; for independent backup lines, the attributes such as backup type and commissioning conditions are marked. Using the path query function of the graph database, the appropriate algorithm is selected according to specific business needs, such as the Dijkstra algorithm to solve the shortest path problem, the A* algorithm to solve the weighted shortest path problem, and the depth-first search (DFS) algorithm to solve the full path problem. When applying the algorithm, set appropriate weights, constraints and other parameters, analyze the key nodes and key paths in the power grid topology, evaluate the importance and impact range of the nodes, and generate network characteristic parameters such as power supply paths and load distribution based on the topology and equipment attributes. For the equipment nodes in the graph structure, associate the ledger attribute information, such as equipment type, rated parameters, manufacturer, etc., to enrich the attribute dimensions of the nodes. For standby equipment, it is also necessary to associate its commissioning conditions, switching methods and other attributes to form a complete standby equipment attribute portrait. Commissioning conditions can include fault types, such as busbar faults, line faults, etc., fault locations, such as XX substations, XX lines, etc., load levels, such as XX% of rated load, etc.; switching methods can include automatic switching, such as through relay protection actions, manual switching, such as through remote operation by dispatchers, etc. These attributes can be designed as enumeration types, numerical types, string types, etc. according to actual needs. After completing the graph structure construction and attribute association, the graph structure is integrated with the geographic information system (GIS). First, unique geographic coordinates and spatial geometric objects are assigned to each device in the GIS system, and then these geographic attributes are associated with the device nodes in the graph structure to achieve the fusion of spatial data and topological data. Based on the integrated graph structure, graph visualization technology is used to design appropriate map symbols, annotation styles, etc., and generate annotation documents containing information such as device nodes, topological connections, and alternate annotations. Different types of devices and connections are distinguished by visual attributes such as node color and shape, and legends are provided to facilitate users to intuitively identify and locate devices. The generated benchmark power grid graph data model is continuously maintained and updated, and the device status and topological structure are dynamically adjusted according to the actual operation of the power grid to ensure that the graph database is always consistent with the site. In the Neo4j graph database, the CREATE statement is used to create device nodes and topological relationship edges. The attributes of device nodes include device ID, device type, rated voltage, etc., and the attributes of topological relationship edges include parameters such as wire material, wire length, and resistance.For backup equipment, the additional_type attribute is marked as "backup", and for backup lines, the backup_line_id attribute is marked with the main line ID. Use the Dijkstra shortest path algorithm to analyze the key power supply path, specify the substation node as the starting point, the load node as the end point, and the line impedance as the weight to calculate the shortest power supply distance. Use the DFS algorithm to traverse the electrical island in the power flow calculation, define the traversal depth as 3, and mark all nodes and edges in the electrical island. Use the NEO4J-SPATIAL plug-in to associate the device node with the geographic coordinates, and use SVG graphics for the substation to define. <circle>Tags, use polyline segments for transmission lines, define<polyl ine> Tags. Use the WGS84 coordinate system in the GIS system and query other devices within 500 meters of the device through spatial index. You can also use Neo4j's version management plug-in neo4j-versioner-core to record the change history of the graph structure, define the attribute name, old value, and new value of the modification operation, record the change timestamp for each version, and support version backtracking.

[0025] S103. Calculate the redundancy of each node in the distribution network based on the backup line and backup equipment data in the benchmark power grid structure and the number of alternative paths between different devices and lines, and generate an electrical topology diagram containing redundancy information and a redundancy information document containing redundancy calculation results.

[0026] The benchmark power grid structure is analyzed, the connection relationship between each node in the distribution network is extracted, and an adjacency matrix representing the network topology is constructed; all reachable nodes of the adjacency matrix are traversed, and the path from the starting node to each reachable node is recorded to obtain a set of candidate paths between all nodes in the distribution network, and each path in the set of candidate paths represents a potential power supply scheme; according to the attribute information of the backup lines and backup equipment, the candidate paths are screened and sorted, and the paths containing the backup lines and backup equipment are preferentially selected, and the paths corresponding to the faulty equipment and the faulty lines are excluded to obtain a set of valid candidate paths for each node; the redundancy index of each node is calculated by combining the set of valid candidate paths and the benchmark power grid structure, and the redundancy index represents The power supply reliability of the node; according to the calculation results of the redundancy index, the distribution network nodes are classified and sorted to generate a distribution network redundancy level classification table; the calculation results of the node redundancy index are associated with the geographical topology information of the distribution network, and the nodes of different redundancy levels are marked with different colors or patterns to generate an electrical topology map integrating the redundancy information to show the power supply reliability distribution of the distribution network; various data and results in the node redundancy analysis process are summarized, including the node adjacency matrix, the set of alternative paths, the set of valid alternative paths, the node redundancy index, the level of redundancy classification, and the distribution network redundancy level classification table to generate an electrical topology map containing redundancy information and a redundancy information document containing the redundancy calculation results.

[0027] Specifically, by analyzing the benchmark power grid structure in the graph database, the connection relationship between each node in the distribution network is extracted, and an adjacency matrix is ​​constructed to represent the network topology. The elements in the adjacency matrix represent the connectivity between nodes. If there is a line connection between two nodes, the corresponding matrix element value is 1, otherwise it is 0. The adjacency matrix is ​​traversed using the depth-first search (DFS) algorithm. Starting from each node, all reachable nodes are explored, and the path from the starting node to each reachable node is recorded. The DFS algorithm uses a stack structure to implement graph traversal. The specific steps include: (1) Select the starting node, mark it as visited, and push it into the stack; (2) When the stack is not empty, pop the top node of the stack and check all its unvisited adjacent nodes; (3) For each unvisited adjacent node, mark it as visited and push it into the stack, and record its connection relationship with the current node; (4) Repeat steps (2) and (3) until the stack is empty. Through the DFS algorithm, a set of alternative paths between all nodes in the distribution network is obtained, and each path represents a potential power supply solution. The set of alternative paths is processed, and the alternative paths are screened and sorted according to the attribute information of the alternative lines and equipment. The paths containing the alternative lines and equipment are given priority, and the paths corresponding to the faulty equipment and lines are excluded, and finally the set of valid alternative paths for each node is obtained. In order to comprehensively consider the power supply capacity and reliability of the alternative paths, the concept of weighted graph is introduced to weight the lines and equipment on the alternative paths. The weight can reflect the power supply capacity and reliability. For example, the weight of the line can be set to its current capacity, and the weight of the equipment can be set to its reliability index. When calculating the redundancy index, not only the number of valid alternative paths is considered, but also the weighted sum of the alternative paths. The calculation formula of the redundancy index is: node redundancy = (weighted sum of valid alternative paths) / (weighted sum of total connected lines). The higher the redundancy index, the better the power supply reliability of the node. According to the calculation results of the redundancy index, refer to industry standards and expert experience, set a scientific redundancy index threshold, and classify and sort the distribution network nodes. According to the order of redundancy index from high to low, it is divided into high redundancy (redundancy index>=0.8), medium redundancy (0.5<=redundancy index

[0028] <0.8) and low redundancy (redundancy index <0.5), and obtain the redundancy level classification table of the distribution network. The nodes of each level are marked and counted to generate a redundancy distribution map or redundancy statistical report. The node redundancy information is associated with the geographical topological information of the distribution network. The nodes of different redundancy levels are marked with different colors or patterns in the electrical topology map to generate an electrical topology map that integrates redundancy information and intuitively displays the distribution of power supply reliability of the distribution network. The Pandas and Matplotlib libraries of Python are used to organize, count and visualize the electrical topology map and other analysis results, design a reasonable document structure and layout, and generate an electrical topology map containing redundancy information and a redundancy information document containing redundancy calculation results. The redundancy information document includes basic information of the distribution network, statistics of backup lines and backup equipment, node redundancy analysis, electrical topology map, conclusions and suggestions, and summarizes the data and results in the node redundancy analysis process. The adjacency matrix of the distribution network is extracted from the graph database. The matrix dimension is 1000*1000, indicating that there are 1000 nodes in the distribution network. Use the NetworkX library to construct a graph object G, and use the G.add_edges_from() method to convert the adjacency matrix into a graph edge set. Use the nx.dfs_tree(G, source = node) function to perform a depth-first traversal of the graph, where node represents the starting node and the traversal result is returned in the form of a tree.

[0029] The nx.all_simple_paths(G,source=node,target=target) function obtains all simple paths from the starting node to the target node to form a set of candidate paths. Assuming that the number of candidate paths is 500, the Pandas DataFrame object is used to store the candidate paths. The column names are "path", "weight", and "rel iabi l ity", which represent the path sequence, power supply capacity, and reliability, respectively. The power supply capacity is calculated by weight according to the current capacity of the line on the path, and the value range is [0,1]; the reliability is calculated by weight according to the reliability index of the device on the path, and the value range is [0,1]. The product of the path weight and reliability is used as the sorting basis, and the top 100 are selected as high-quality candidate paths. The redundancy index is calculated for each node, and the mean of the redundancy index is 0.75 and the standard deviation is 0.2. The Matplot lib library is used to draw a distribution histogram of the redundancy index, with the horizontal axis being the redundancy index and the vertical axis being the number of nodes. The redundancy index thresholds were set to 0.8 and 0.5, and 200 high-redundancy nodes, 500 medium-redundancy nodes, and 300 low-redundancy nodes were obtained. The geographic coordinate information of the distribution network nodes was combined with the redundancy index using the Geopandas library to generate a redundancy geographic topology map of the distribution network. The redundancy analysis report was generated using the Pandas DataFrame object and Matplotlib library and exported to HTML format. The redundancy information document contains 4 chapters, namely basic information of the distribution network, alternative path analysis, redundancy index distribution, and redundancy level classification. Each chapter contains 1-2 charts and corresponding analysis text.

[0030] S104. Based on the redundancy index calculated by the node, analyze the fault recovery capability of the distribution network at different redundancy levels, evaluate its switching and recovery speed when a fault occurs, and generate a fault analysis document including the fault recovery capability analysis results and the switching and recovery speed evaluation results when a fault occurs.

[0031] Based on the redundancy level classification table of the distribution network, a fault simulation analysis is performed on each redundancy level, including calculating the power supply recovery time of the node in the event of a fault according to the number of alternative paths and the power supply capacity of the node, and obtaining the fault recovery time data at different redundancy levels; analyzing the influence of the automatic switching mode and the manual switching mode on the fault switching time, and obtaining the fault switching time data; using the fault recovery time data and the fault switching time data, the fault recovery capability of the distribution network at different redundancy levels is evaluated, including calculating the fault recovery capability index; analyzing the correlation between the fault recovery capability index and the redundancy index, fitting the functional relationship between the redundancy index and the fault recovery capability index, obtaining the influence of the redundancy index on the fault recovery capability, and forming a quantitative relationship model between the redundancy index and the fault recovery capability; summarizing the evaluation results of the fault recovery capability of the distribution network at different redundancy levels and the quantitative relationship model between the redundancy index and the fault recovery capability, and generating a fault analysis document containing the fault recovery capability analysis results and the evaluation results of the switching and recovery speed when a fault occurs.

[0032] Specifically, based on the redundancy level classification table of the distribution network, nodes with different redundancy levels have different recovery capabilities and switching speeds when a fault occurs. For each redundancy level, the PSASP power system simulation software is used to perform fault simulation analysis. According to the distribution network topology and equipment parameters, a simulation model is constructed, and the fault scenario and simulation parameters are set to simulate the situation where the node fails. According to the number of alternative paths and power supply capacity of the node, the power supply recovery time of the node under fault conditions is calculated, and the average fault recovery time data under different redundancy levels are obtained to form a corresponding relationship table between the redundancy level and the fault recovery time of the distribution network. On the basis of the fault simulation analysis, the fault switching mode of the node is further considered, and the automatic switching and manual switching are distinguished. Automatic switching is achieved through intelligent switches and fault location technology, and manual switching requires manual operation. According to different switching modes, the fault switching time of the node is calculated, and the average fault switching time data under different redundancy levels are obtained. Using the fault recovery time and fault switching time data, the fault recovery capability of the distribution network at different redundancy levels is comprehensively evaluated. The standard recovery time is defined as a reference recovery time determined according to industry standards or empirical values, indicating that power restoration within this time is considered acceptable. Calculate the fault recovery capability index. The higher the index is, the stronger the fault recovery capability of the distribution network is at this redundancy level. The calculation formula of the fault recovery capability index is: fault recovery capability index = (standard recovery time - average fault recovery time) / standard recovery time × 100%. On the basis of calculating the fault recovery capability index, further analyze the correlation between the fault recovery capability index and the redundancy index. Draw a scatter plot of the redundancy index and the fault recovery capability index to observe the correlation between the two. According to the shape of the scatter plot, select a suitable regression model, such as linear regression, polynomial regression, exponential regression, etc. Use the least squares method to estimate the model parameters and obtain the coefficients and intercept of the regression model. Use the coefficient of determination (R 2 ), root mean square error (RMSE) and other indicators are used to evaluate the goodness of fit and prediction accuracy of the regression model, and a quantitative relationship model between redundancy indicators and fault recovery capability is obtained. The evaluation results of the fault recovery capability of the distribution network at different redundancy levels and the quantitative relationship model between redundancy indicators and fault recovery capability are summarized, and a fault analysis document containing the fault recovery capability analysis results and the evaluation results of the switching and recovery speed when a fault occurs is generated. In order to realize the real-time monitoring of the redundancy of the distribution network, monitoring devices such as smart meters and miniature circuit breakers are deployed in the distribution network, and the current, voltage, power and other data of each node are collected in real time through wireless communication or power line communication technology. The collected data is cleaned, stored and analyzed, the real-time redundancy indicators of the nodes are calculated, and a redundancy monitoring report and early warning information are formed. According to the results of fault recovery capability analysis and real-time redundancy monitoring, the distribution network is optimized. For the key nodes and important loads in the distribution network, targeted optimization strategies are formulated according to their redundancy indicators and fault recovery capability indicators, such as adding spare lines and improving equipment reliability. By optimizing the topology and equipment configuration of the distribution network, the redundancy level of key nodes and important loads is improved, thereby improving their fault recovery capability. The results of fault recovery capability analysis can also be applied to the planning and operation management of distribution networks. In the distribution network planning stage, the redundancy level of the distribution network is reasonably set according to the load forecast and reliability requirements. In the distribution network operation stage, the operation mode of the distribution network is dynamically adjusted according to the real-time redundancy monitoring data to ensure that the distribution network can maintain good fault recovery capabilities at different redundancy levels. The redundancy data of 500 distribution network nodes are obtained from the distribution automation system. The data includes attributes such as node ID, redundancy index, and power supply capacity. Using the K-means clustering algorithm, the number of clusters k=3 is set to cluster the nodes, and three redundancy levels of high, medium, and low are obtained. The cluster centers are 0.85, 0.60, and 0.35, respectively. A distribution network simulation model is built on the PSASP simulation platform. The model includes 500 nodes, 750 lines, and 100 switches. Three fault scenarios are set: single-point grounding, two-point grounding, and three-phase short circuit. 100 fault points are randomly selected in each fault scenario. Run the simulation program and count the average fault recovery time of each node. The high redundancy level is 20 minutes, the medium redundancy level is 40 minutes, and the low redundancy level is 60 minutes. For nodes that switch automatically, the switching time is calculated as 2 minutes; for nodes that switch manually, the switching time is calculated as 10 minutes. Assuming that the standard recovery time is 30 minutes, the fault recovery capability index of each redundancy level is calculated, which is 93.33%, 83.33%, and 66.67% respectively. Draw a scatter plot of the redundancy index and the fault recovery capability index, and find that the two show a significant positive correlation.We tried linear regression, logarithmic regression, exponential regression and other models, and finally chose the logarithmic regression model, the regression equation is y = 0.2030ln (x) + 0.8616, where x is the redundancy index and y is the fault recovery capability index. The determination coefficient R of the model. 2 =0.8762, RMSE =0.0336, good fit. 100 nodes were selected in the distribution network, smart meters and miniature circuit breakers were installed, and data was uploaded once a minute through ZigBee wireless communication. Hadoop was used to build a data platform, Hive was used for data cleaning and conversion, SparkStreaming was used for real-time calculation, and the redundancy index was updated every 5 minutes. When the redundancy index is lower than 0.4, the system automatically pushes warning information to prompt the operation and maintenance personnel to deal with it in time. For important nodes with redundancy index lower than 0.2, optimization plans were formulated, such as adding a backup line, replacing an old transformer, etc. After optimization, the average redundancy index of important nodes was increased to 0.56. When planning the distribution network, according to the load forecast data and reliability requirements, the redundancy target values ​​of each area were set, such as 0.80 in urban areas and 0.60 in rural areas. When the distribution network is in operation, the real-time redundancy index is evaluated every 15 minutes. When the index is lower than the target value, it automatically switches to the backup line or backup power supply to improve the reliability of power supply.

[0033] S105. Evaluate the resource utilization of the distribution network based on the historical fault data of the distribution network in combination with the redundancy index, determine the idle status of the backup equipment and lines in a non-fault state, and generate an evaluation document containing the historical fault data and the resource utilization evaluation results.

[0034] Obtain historical fault records of the distribution network, wherein the fault records include fault occurrence time, faulty equipment, fault type, impact range, and repair time information; preprocess the historical fault records of the distribution network; form a distribution network fault data set by summarizing the preprocessed historical fault records of the distribution network; spatially map the distribution network fault data set according to the topological structure and equipment ledger information of the distribution network to determine the specific location of each fault event; associate the specific location with the attribute information of the faulty equipment to form an attribute portrait of the faulty equipment; use the attribute portrait of the faulty equipment to analyze the fault characteristics and laws of different types of equipment; perform correlation analysis on the distribution network fault data set based on the redundancy information document, and calculate each redundancy The failure rate, failure duration, and failure impact range indicators at different levels are used to evaluate the impact of the redundancy level on the reliability of the distribution network; a quantitative relationship between redundancy and reliability is established based on the evaluation results; the relationship between faults, types, and redundancy is analyzed to find out the main causes and risk factors of faults; the utilization efficiency of backup resources is evaluated to identify the optimal balance between redundancy and resource utilization; a fault prediction model is constructed to predict the probability and type of equipment failures; the analysis results of the failure characteristics and laws of different types of equipment, the quantitative relationship between redundancy and reliability, the main causes and risk factors of failures, the evaluation results of the utilization efficiency of backup resources, and the failure prediction results are summarized to generate an evaluation report on the distribution network's historical fault data analysis and resource utilization results.

[0035] Specifically, the fault records of the distribution network in the past year are extracted from the historical database of the distribution automation system, including information such as the time of fault occurrence, faulty equipment, fault type, impact range, and repair time, to form a distribution network fault data set. The fault data is cleaned and preprocessed, incomplete and inconsistent data are eliminated, and the fault data is standardized into a structured data format. According to the topological structure and equipment ledger information of the distribution network, the fault data is spatially mapped to determine the specific location of each fault event and associate the attribute information of the faulty equipment, such as equipment type, rated parameters, manufacturer, etc. Key attributes such as equipment type, fault type, fault frequency, and fault duration are selected, and the attribute values ​​are discretized or normalized to form a standardized attribute feature vector. Then, visualization methods such as radar charts and parallel coordinate charts are used to intuitively display the characteristic distribution of faulty equipment on different attributes to form an attribute portrait of the faulty equipment. Combined with the redundancy information of the distribution network, the fault data is correlated and analyzed, and indicators such as the failure rate, failure duration, and failure impact range under each redundancy level are calculated to evaluate the impact of the redundancy level on the reliability of the distribution network. Statistical methods, such as chi-square test and correlation coefficient, are used to analyze the correlation between fault data and factors such as equipment attributes and redundancy, and to find out the key factors that affect the occurrence of faults. The quantitative relationship between redundancy and reliability is obtained. Association rule mining algorithms, such as Apriori algorithm or FP-growth algorithm, are used to deeply mine fault data and redundancy information. The minimum support is set to 10% and the minimum confidence is set to 60%, and frequent item sets and association rules that meet the support and confidence thresholds are mined. Then, indicators such as lift and chi-square value are used to evaluate the effectiveness and significance of association rules and screen out strong association rules. The influence of different factors on the occurrence of faults is evaluated, and the main causes and risk factors of faults are obtained. Based on the analysis of fault data, the utilization of spare equipment and spare lines is evaluated. The idle time and idle frequency of spare equipment and spare lines in non-fault state are counted, and the average utilization rate and maximum idle time of spare resources are calculated. The utilization of spare resources under different redundancy levels is compared to obtain the balance point between redundancy and resource utilization. Select commonly used classification algorithms such as decision trees, random forests, and support vector machines to establish a fault prediction model. Input features such as equipment attributes, redundancy, and environmental factors to perform parameter tuning and cross-validation on the model. Use evaluation indicators such as precision, recall, F1 value, and ROC curve to judge the classification performance of the model. At the same time, use methods such as feature importance sorting and feature selection to optimize the feature space and generalization ability of the model. Summarize the results of distribution network fault data analysis, reliability assessment, association rule mining, backup resource utilization assessment, and fault prediction, and generate evaluation documents for distribution network historical fault data analysis and resource utilization assessment.The document content includes fault data statistics, redundancy impact analysis, association rule mining results, backup resource utilization evaluation and fault prediction model evaluation; the fault data statistics part includes the time distribution, regional distribution, equipment type distribution, etc. of fault occurrence; the redundancy impact analysis part includes the comparison of failure rate and failure duration under different redundancy levels, and the correlation analysis of redundancy and reliability indicators; the association rule mining results part includes frequent item sets and association rule lists, and the association network diagram of key influencing factors; the backup resource utilization evaluation part includes the average utilization rate of backup equipment and backup lines, the maximum idle time statistics, and the balance analysis of redundancy and resource utilization; the fault prediction model evaluation part includes the classification performance indicators of the model, feature importance ranking, and the prediction results of typical fault cases. From the Oracle database of the distribution automation system, through SQL query statements, a total of 2000 fault records from January 1, 2022 to December 31, 2022 are extracted. The fault data is cleaned using Python's Pandas library, and fields with missing values ​​exceeding 20% ​​are removed, and fields with missing values ​​less than 20% are filled with mean or mode. Regular expressions are used to standardize the names of faulty equipment, and different types of circuit breakers, transformers and other equipment are normalized into a unified equipment type. Baidu Map API is used to convert the latitude and longitude coordinates of the faulty equipment into specific geographic location information. The fault frequency and fault duration are normalized to the minimum-maximum value, and the attribute values ​​are mapped to the [0,1] interval. A scatter plot of the faulty equipment on the two attributes of fault frequency and fault duration is drawn, and the equipment types are distinguished by different colors. The Apriori algorithm is used to mine association rules for fault data and redundancy information, setting the minimum support to 0.1 and the minimum confidence to 0.6, and a total of 50 association rules are mined. The lift and chi-square value of each rule are calculated, and the rules with a lift greater than 2 and a chi-square value greater than 3.841, with a degree of freedom of 1 and a chi-square critical value of 0.05 are selected as strong association rules, and 10 strong association rules are obtained. The proportion of idle time of standby equipment and standby lines to the total time is counted. The average idle rate of standby transformers is 35%, and the average idle rate of standby lines is 40%. The fault prediction model was established using the random forest algorithm. Ten features, including fault frequency, fault duration, equipment type, and redundancy, were selected as inputs, and the fault type was used as output. 70% of the data was divided into a training set and 30% of the data was used as a test set. Through grid search and 5-fold cross validation, the optimal parameters of the random forest were determined to be 100 trees, 10 maximum tree depth, and 5 minimum node samples. The model performance was evaluated on the test set, with an accuracy of 85%, a precision of 82%, a recall of 87%, an F1 value of 0.84, and an AUC value of 0.91.Generate an evaluation document for fault data analysis, including: an overview of the fault data of distribution network A, where equipment type A has the highest fault frequency, accounting for 35% of the total number of faults; distribution network B has the longest average fault recovery time, which is 55 minutes. As the redundancy increases from 0.2 to 0.8, the failure rate drops from 15% to 3%, and the average fault recovery time is shortened from 60 minutes to 20 minutes. The random forest model has an 85% accuracy rate in predicting fault types on the test set, and the highest accuracy rate is for transformer faults, reaching 92%. It can also include recommendations for key monitoring and maintenance of transformers, a reasonable allocation of the number of spare transformers, and regular inspections and repairs of high-risk lines to ensure safe and stable operation of the distribution network.

[0036] S106. Summarize the inventory documents, annotation documents, redundancy information documents, fault analysis documents and evaluation documents of the power grid equipment, perform preprocessing, and build a vocabulary; represent each document as a word frequency vector of the word in the vocabulary, and generate a document-word frequency matrix, where the rows of the matrix represent documents, the columns represent words, and the values ​​in the matrix represent the frequency of the words appearing in the document.

[0037] Summarize the ledger documents, annotation documents, redundancy information documents, fault analysis documents and evaluation documents of power grid equipment, convert the summarized document contents into plain text format, and remove non-text elements, including pictures and tables, to obtain a plain text data set; preprocess the plain text data set, and the preprocessing includes word segmentation, stop word removal, and part-of-speech tagging operations; perform part-of-speech tagging on the preprocessed plain text data set to form a vocabulary; count the word frequency based on the vocabulary, select words with a frequency exceeding a threshold as keywords, and form a keyword list; traverse each document, count the frequency of each word in the keyword list in the document, and form a word frequency vector of the document, the dimension of the word frequency vector is equal to the length of the keyword list, and each element of the word frequency vector represents the frequency of the corresponding keyword in the document; construct a document-word frequency matrix based on the word frequency vectors of each document, the rows of the matrix represent documents, the columns represent words, and the values ​​in the matrix represent the frequency of words in the document.

[0038] Specifically, the ledger documents, annotation documents, redundancy information documents, fault analysis documents and evaluation documents of power grid equipment are summarized from different sources, including ledger attributes and electrical topology connection relationships, benchmark power grid structure and backup lines, backup equipment data, redundancy calculation results, fault recovery capability evaluation, historical fault data and resource utilization evaluation results, etc., and the document content is converted into plain text format. Regular expression matching and document parsing tools are used to remove non-text elements such as pictures and tables in text data to extract plain text information. The extracted text data is cleaned and filtered, the Simhash algorithm is used to identify duplicate documents, and regular expression matching is used to identify blank documents and documents with incorrect formats, and they are removed or corrected to obtain the preprocessed text data set. The preprocessed text data set is segmented, stop words are removed, and part-of-speech tagging is performed. The jieba word segmentation tool is used, and the dictionary-based forward maximum matching algorithm is used for word segmentation to remove common words in the stop word list. The stop word list is built based on the Chinese stop word corpus. Use the NLTK library to perform part-of-speech tagging on the word segmentation results, and mark the part of speech of each word. The tag set can be the PKU tag set of the Institute of Computational Linguistics of Peking University. Based on the preprocessed text data, build a vocabulary, count all non-repeated words that appear in the text data, and generate a vocabulary list. Calculate the frequency of occurrence of each word in the document set, divide it by the total number of documents, and get the document frequency (DF) of the word. Calculate the frequency of occurrence of each word in each document, divide it by the length of the document, and get the word frequency (TF) of the word. Calculate the TF-IDF value of the word, sort it from high to low according to the TF-IDF value, select the top-80 words as keywords, and form a keyword list. Represent each document as a frequency vector of the words in the vocabulary, and the dimension of the vector is equal to the size of the vocabulary. Traverse each document, count the frequency of each word in the vocabulary in the document, and generate the document's word frequency vector. Each element of the vector represents the frequency of the corresponding word in the document, and the frequency of the word that does not appear is 0. Finally, a document-word frequency matrix is ​​formed. The rows of the matrix represent documents, the columns represent words in the vocabulary, and the elements in the matrix represent the frequency of the word in the document. Considering the high-dimensional sparse characteristics of the document-word frequency matrix, the CSR (Compressed Sparse Row) format is used to compress and store the matrix, storing only the non-zero elements in the matrix and their position information, greatly reducing the storage space of the matrix and improving the efficiency of matrix operations. The document data from 10 business systems such as the equipment ledger system, operation and maintenance management system, and fault diagnosis system are summarized, totaling 1,000 documents and 1GB of data. Use regular expressions"<img.?> |<table.?> |<chart.*?> "Match and remove non-text elements such as pictures, tables, and charts in the text, with a removal rate of 95%. Use the Simhash algorithm to detect the duplication of documents, and remove duplicate documents with a Hamming distance threshold of 5 as the standard, with a removal rate of 8%.Use the jieba word segmentation tool to segment the text, and use the Viterbi algorithm to perform Chinese word segmentation based on HMM, with an average segmentation accuracy of 96%. The part-of-speech tagging uses the PennTreebank tagging set of the NLTK library, with a tagging accuracy of 94%. Construct a vocabulary that contains 20,000 non-repeating words. Calculate the TF-IDF value of each word and take the top-100 words as keywords. Generate a 10000×20000 document-word frequency matrix and use CSR format for compression storage, with a compression rate of .

[0039] S107, training the LDA model, setting the number of topics, performing iterative training, and extracting the topic distribution of each document and the vocabulary distribution of each topic.

[0040] According to the document-word frequency matrix, the input of the LDA topic model is constructed, and each document is represented as a frequency vector or a weighted vector of the vocabulary and is normalized to ensure that the sum of the vector elements is 1; the number of topics of the LDA model is set in combination with the document characteristics and domain knowledge; the parameters of the LDA model are initialized, including the topic-word distribution parameters and the document-topic distribution parameters; the LDA model is trained to gradually converge to the optimal topic-word distribution and document-topic distribution; the posterior probability of each topic is calculated, and the topic-word distribution and document-topic distribution are updated; based on the updated topic-word distribution and document-topic distribution, the vocabulary distribution of each topic and the topic distribution of each document are obtained.

[0041] Specifically, according to the document-word frequency matrix, the input format of the LDA topic model is constructed, each document is represented as a frequency vector or weight vector of the vocabulary, and normalized to ensure that the sum of the vector elements is 1, satisfying the probability distribution constraint of the LDA model. The specific format of the document-word matrix can be a TF matrix, a TF-IDF matrix or a word frequency matrix. By counting the frequency or weight of each word in each document, a sparse matrix representation is generated. The number of topics of the LDA model is set. According to the scale of the document set and domain knowledge, the appropriate number of topics is selected, usually in the range of 10ˉ100. The optimal number of topics can be determined by cross-validation, perplexity evaluation, topic consistency evaluation and other methods to balance the complexity and explanatory power of the model. Cross-validation selects the number of topics with the best generalization performance by calculating the prediction error of the LDA model on the test set; perplexity evaluation selects the number of topics with the smallest perplexity by calculating the probability prediction ability of the LDA model on the test set; topic consistency evaluation selects the number of topics with the highest topic consistency by calculating the semantic similarity between topics. Initialize the parameters of the LDA model, including the topic-word distribution parameter β and the document-topic distribution parameter θ. You can use the symmetric Dirichlet prior distribution for random initialization, set the prior parameters α and η, and control the shape and scale of the prior distribution. The common values ​​are α=50 / K, η=0.1 or 0.01, where K is the number of topics. You can also draw a perplexity curve, that is, take the number of topics K as the horizontal axis and the corresponding perplexity value as the vertical axis, and determine a suitable K value by observing the trend of the curve. The ideal K value is usually near the inflection point of the perplexity curve, that is, the position where the perplexity decreases slowly. Hyperparameters α and β have an important influence on model training. α controls the prior of document-topic distribution, and β controls the prior of topic-word distribution. A larger α will make each document tend to contain more topics, and a larger β will make each topic tend to contain more words. Hyperparameter optimization can be performed through grid search and other methods. The Gibbs sampling algorithm is used to train and infer the LDA model. By iteratively updating the topic assignment of each word in the document, it gradually converges to the optimal topic-word distribution and document-topic distribution. Gibbs sampling approaches the posterior distribution by iteratively sampling the topic assignment of each word. The number of iterations is usually set to 1000-2000 times, and each iteration needs to traverse all the words in the document set. For each document, traverse each word position in the document, calculate the posterior probability of each topic based on the topic assignments of other words and the current model parameters, and perform topic sampling and updating. During the sampling process, optimization techniques such as folded Gibbs sampling and topic block sampling can be used to improve sampling efficiency and convergence speed. In each iteration of Gibbs sampling, the posterior probability of each topic is calculated based on the topic assignment of the current word and the topic assignments of other words, and topic sampling is performed based on the posterior probability, the topic assignment of the current word is updated, and the corresponding topic-word distribution and document-topic distribution are updated.After the iteration, the parameters of the LDA model are estimated according to the obtained topic-word distribution and document-topic distribution, and the vocabulary distribution of each topic and the topic distribution of each document are obtained. The probability distribution can be obtained by normalizing the frequency matrix, or the sampling results of Gibbs sampling can be used to calculate the posterior expectation of the word distribution of each topic and the topic distribution of each document as the parameter estimation value. The trained LDA model is used to perform topic inference and classification on new documents. The topic distribution of the document is obtained by calculating the posterior probability of the document on each topic, and the topic with the highest probability is selected as the topic label of the document. The document-word frequency matrix of 3000 devices, 5000 lines, and 1000 topic words is obtained by statistics using documents such as power grid equipment ledgers, topological structures, spare lines, and fault data. The optimal number of topics is determined to be 50 through topic consistency evaluation and perplexity curve. The topic-word distribution parameter β and the document-topic distribution parameter θ are initialized, and the prior parameters α=1 and η=0.1 are set. The LDA model is trained using the folded Gibbs sampling algorithm, iterated 1500 times, and each iteration takes 10 seconds. The sampling process was parallelized using OpenMP multithreading to speed up the training time. The training obtained a 50×1000 topic-word distribution matrix and a 3000×50 document-topic distribution matrix, and the coverage of the top 10 topic words reached 85%. 500 new equipment ledger documents were selected for topic inference, with an average inference time of 20ms / document. The topic distribution features of each document were extracted and fused with TF-IDF features to train the support vector machine classification model, achieving an accuracy of 90% in the equipment status classification task. The LDA model was applied to the topic clustering of distribution equipment, and 10 typical equipment combination patterns were found, such as "oil-immersed transformer-vacuum circuit breaker-cable", which matched 90% of the equipment combinations in the expert knowledge base. The Word2Vec word embedding representation was introduced to map the topic-word distribution to a low-dimensional semantic space. The t-SNE algorithm was used to visualize the topic distribution, clearly present the semantic structure of distribution professional vocabulary, and discover key topic clusters such as "fault warning" and "status maintenance".

[0042] S108. Based on the redundancy information, a feasibility report generation model is constructed. The feasibility report generation model is used to analyze potential themes and patterns in the redundancy information and generate a distribution network redundancy feasibility report.

[0043] Based on the redundancy information of each link in the distribution network, the LDA topic model is trained to obtain the indicator distribution of each topic and the topic distribution of each device or node, and the potential semantic structure and association pattern in the redundancy information are mined; the topic-indicator distribution is analyzed, each topic is interpreted and named, the redundancy pattern and characteristics represented by the topic are summarized, the rationality and practical significance of each topic are evaluated, and representative redundancy topics are screened out; using the topic distribution of equipment or nodes, equipment or nodes with similar redundancy characteristics are classified into the same category to form a classification of redundant equipment or nodes, and the redundancy level and risk degree of each category are analyzed; the redundancy pattern mined by the LDA topic model and the equipment / node classification results are integrated to construct a redundancy feasibility report generation model, and generate a corresponding feasibility analysis report according to the topic distribution characteristics of the equipment or node.

[0044] Specifically, the redundancy information of each link of the distribution network can also include the redundancy indicators of each device and node such as power supply, substation, line, switch, load, etc. The data source can include power monitoring system, equipment management system, fault diagnosis system, etc. The data format can include CSV, JSON, Excel, etc. The redundancy information of each link of the distribution network is cleaned, standardized and integrated, and the problems of missing values, outliers, inconsistency, etc. are handled. A unified data format and representation method are constructed to form a standardized redundancy data set to prepare data for subsequent topic mining and pattern analysis. Exploratory analysis is performed on the standardized redundancy data set. The pandas library is used to read the data. The describe() function is used to calculate the statistical characteristics of each redundancy indicator, such as mean, variance, quantile, etc. The groupby() function is used for group analysis. The matplotlib library is used to draw the indicator distribution histogram and box plot. The seaborn library is used to draw the correlation heat map, etc. The outliers and outliers in the data are identified through visual analysis, the distribution characteristics and correlation patterns of the data are discovered, and the abnormal data are checked and processed to ensure data quality and consistency. The redundancy dataset is converted into a document-term frequency matrix or TF-IDF matrix. Each document corresponds to a device or node. The vocabulary consists of redundancy indicators. The elements in the matrix represent the values ​​or weights of the indicators on the device or node, and the input representation suitable for the LDA model is constructed. Based on the converted redundancy data, the LDA topic model is trained. In order to optimize the model performance, multiple evaluation indicators such as perplexity, coherence score, and topic distinctiveness are used to quantify the model effect. By performing strategies such as grid search, the key hyperparameters of the LDA model are systematically adjusted. These hyperparameters include the number of topics, alpha (controlling the smoothness of the word distribution within a topic), and beta (controlling the smoothness of the topic distribution between documents). The goal is to find a combination of hyperparameters that achieves the best balance of evaluation indicators, so as to ensure that the model can accurately capture the topic structure in the data and show good explanatory power. Use methods such as 5-fold cross validation to evaluate the generalization performance of the model, obtain the indicator distribution of each topic and the topic distribution of each device or node, and mine the potential semantic structure and association pattern in the redundant information. Analyze the topic-indicator distribution obtained by the LDA model, explain and name each topic, use TF-IDF weighted topic word sorting, extract keywords for each topic as topic labels, or use word embedding and other methods to calculate the semantic similarity of topic words and cluster to generate topic labels.Combined with domain knowledge and expert experience, summarize the redundancy patterns and features represented by each topic, such as "high redundancy topics" and "low redundancy topics", evaluate the rationality and practical significance of each topic, and screen out important and representative redundancy topics. Use the topic distribution of devices or nodes obtained by the LDA model to cluster and group devices or nodes, and classify devices or nodes with similar redundancy characteristics into the same category to form a classification system for redundant devices or nodes. Analyze the redundancy level and risk level of each category, identify key redundancy weaknesses and potential improvement opportunities. Combine the redundancy patterns mined by the LDA topic model, the equipment / node classification results, and the domain expert knowledge to build a redundancy feasibility report generation model. Use a template-based natural language generation method to predefine the framework and chapter structure of the feasibility report, such as "current situation assessment", "problem analysis", "improvement suggestions", etc. According to the topic distribution characteristics of the device or node, fill in the corresponding content template to generate a complete feasibility report. You can also use deep learning text generation models, such as GPT, BERT, etc., to fine-tune on large-scale report corpora to generate more flexible and personalized feasibility reports. The redundancy data of the distribution network at two voltage levels of 10kV and 35kV can be extracted from six business systems such as the distribution automation system and the equipment management system, including the redundancy indicators of 110 main transformers, 326 feeders, 584 switches, and 1320 ring main units, forming a redundancy data set of 2GB and 15,000 records. The Pandas library of Python is used to clean the data, remove the fields with missing values ​​exceeding 20%, perform nearest neighbor interpolation filling for the fields with missing values ​​below 20%, perform maximum-minimum normalization on numerical indicators, and perform one-hot encoding on enumeration indicators to form a standardized data representation. Matplotlib is used to draw the frequency distribution histogram of the redundancy indicators, and truncate the outliers below the 5% quantile and above the 95% quantile. The TF-IDF algorithm is used to convert the redundancy data into a 1320×50 document-term frequency matrix, where each row represents a ring main unit and each column represents a TF-IDF weighted redundancy indicator. The Gens im library was used to train the LDA topic model, with the number of topics set to 10, α = 0.1, β = 0.01, and the number of iterations set to 1000. CoherenceScore was used as the evaluation indicator, and the average CoherenceScore obtained by 5-fold cross validation was 0.76. The top five keywords of each topic were extracted, such as "capacity margin", "power redundancy", "switch configuration", "number of line cores", "wiring method", etc., and the topics were manually named as "high redundancy", "medium redundancy", "low redundancy", "N-1 redundancy", "N-2 redundancy", etc. The t-SNE algorithm was used to reduce the dimension of the document topic distribution and visualize it. It was found that the ring main unit with the "high redundancy" topic was clustered in the 2D space, while the ring main unit with the "low redundancy" topic was dispersed.The K-Means algorithm is used to cluster the ring main units, with the number of clusters being 5. The numbers of ring main units at each redundancy level are 205, 326, 418, 256, and 115, respectively. The Jinja2 template engine is used to automatically generate a feasibility analysis report, which includes "Overall redundancy level assessment of the distribution network: The average redundancy index of the distribution network is 0.72, which is at a medium level. The redundancy of the 35kV power grid is better than that of the 10kV power grid, and the redundancy of the rural power grid is lower than that of the urban power grid"; it can also include "Weak link analysis: The redundancy level of the ring network cabinets numbered #254 and #380 is the lowest, and there are weak links such as "single line" and "missing switches". It is recommended to increase spare lines and add contact switches"; it can also include "Improvement measures: For the 110 main transformers on the power supply side, increase the capacity margin and expand the spare capacity; for the 326 feeders on the line side, increase the number of line cores and increase spare lines; for the 584 switches on the switch side, optimize the jumper mode and increase bypass backup". The generated feasibility analysis report has been reviewed by experts with an accuracy rate of 95%, which can be used to guide the planning and construction of the distribution network and improve the reliability of the distribution network power supply.

[0045] The feasibility report generation model combines with the local client or database to generate a distribution network redundancy feasibility report. The distribution network redundancy feasibility report includes redundancy calculation results, resource utilization analysis and feasibility evaluation, and performs rule verification.

[0046] Read data including equipment inventory, topological structure, and redundancy calculation results of the distribution network; call the feasibility report generation model, fill in the predefined report template, and generate a first draft of the distribution network redundancy analysis report; associate and integrate the generated first draft of the distribution network redundancy analysis report with the distribution network resource utilization analysis results, evaluate the balance between the distribution network redundancy level and resource utilization efficiency, and form a feasibility analysis conclusion; integrate the generated first draft of the distribution network redundancy analysis report with the feasibility analysis conclusion; perform rule verification on the integrated distribution network redundancy feasibility analysis report; optimize and polish the distribution network redundancy feasibility analysis report after rule verification; digitally compile the optimized and polished distribution network redundancy feasibility analysis report; check and approve the compiled distribution network redundancy feasibility analysis report to form a final distribution network redundancy feasibility analysis report.

[0047] Specifically, a data connection is established between the local client and the database, and data interaction and synchronization between the client and the database are realized through database connection technologies such as JDBC or ODBC. The database adopts relational databases such as MySQL, Oracle, etc., and designs equipment tables, line tables, node tables, etc., and uses foreign key associations to represent the topological connection relationship between devices. The client reads basic data such as equipment ledgers, topological structures, and redundancy calculation results of the distribution network from the database to ensure data consistency and real-time performance. In the local client, based on the basic data of the distribution network obtained from the database, the feasibility report generation model is called, and the feasibility report generation model is developed using languages ​​such as Java and Python based on document template engines such as Freemarker and Velocity. By reading the basic data of the distribution network and filling in the predefined report template, the first draft of the distribution network redundancy analysis report is generated. In the local client, the report generation model is called through the API interface or command line. The report content includes the overall evaluation of the redundancy of the distribution network, the analysis of weak links, potential improvement opportunities, etc., forming the basic framework and content of the report. The automatically generated draft of the distribution network redundancy analysis report is associated and integrated with the results of the distribution network resource utilization analysis. Using data warehouse technology, the redundancy analysis results and resource utilization analysis results are loaded into the same data mart or data set. Then, SQL statements or OLAP tools are used to perform data association queries and aggregation analysis to evaluate the balance between the redundancy level of the distribution network and the resource utilization efficiency, and form a comprehensive feasibility analysis conclusion. The integrated distribution network redundancy feasibility analysis report is subjected to comprehensive rule verification and quality inspection. According to the distribution network planning and design specifications and standards, such as "Technical Regulations for Power Planning and Design" (DL / T5729-2016) and "Technical Guidelines for 20kVˉ110kV Distribution Network Planning and Design" (Q / GDW1738-2012), these specifications and standards are converted into executable verification rules and embedded into the feasibility report generation model. Compliance checks are performed on the redundancy indicators and resource utilization indicators in the report, and errors, omissions, inconsistencies, etc. in the report are identified, prompting users to make corrections and improvements. On the basis of rule verification, the feasibility analysis report of redundancy of distribution network is further optimized and polished. Natural language processing toolkits such as StanfordCoreNLP and NLTK are used to process the report text through word segmentation, part-of-speech tagging, syntactic analysis, etc. Then, text summarization and text rewriting are used to optimize the language expression of the report. Text generation models such as seq2seq and Transformer can be used in combination with the professional corpus of distribution network to transform and polish the semantics of the report, improve the readability and professionalism of the report, and generate high-quality reports that meet industry standards and user needs.The optimized distribution network redundancy feasibility analysis report is compiled and published in a digital form. The generated HTML format report is converted into PDF format using Java libraries such as iText and ApachePOI. By embedding electronic document generation tools such as MicrosoftOffice, WPSOffice, and LaTeX in the local client, it is typeset and organized in the form of printable and interactive electronic documents. Using JavaScript chart libraries such as ECharts and Highcharts, interactive charts and data visualization components are embedded in the HTML report to enhance the readability and interactivity of the report and realize the digital compilation and publication of the report. The compiled distribution network redundancy feasibility analysis report is subject to multi-level verification and approval. The Activiti workflow engine is used to define the process model for report verification and approval, including various approval nodes, approval roles, approval conditions, etc., and the process model is deployed to the workflow system. In the local client, the report verification and approval process is initiated by calling the RESTAPI of the workflow engine to track the execution status and results of the process. Using the JIRA task management system, we create task orders for report verification and approval, assign them to experts and managers in related fields, record the execution process and feedback of the tasks, and verify and confirm the key contents, conclusions, and suggestions in the report to ensure the accuracy, reliability, and executability of the report. Finally, a formal feasibility analysis report on the redundancy of the distribution network is formed to provide decision support for distribution network planning, construction, operation and maintenance activities. Using the MySQL relational database, we design the distribution network resource data table, including 10 data tables such as equipment table, line table, and node table, with a total of 120 fields and a data volume of 10GB. Using the JDBC connection pool technology, we set the maximum number of connections to 50 and the connection timeout to 5 seconds to achieve efficient connection between the client and the database. Based on the Python FlaskWeb framework and the Jinja2 template engine, we develop a distribution network redundancy analysis report generation service and define more than 10 report templates, covering multiple aspects such as overall evaluation of the distribution network, regional evaluation, and weak link analysis. Using the Pandas data analysis library, we performed correlation analysis on the redundancy analysis results and resource utilization data, generated more than 20 pivot tables and cross tables, and used the Matplotlib visualization library to generate 10 different types of charts. We applied the long short-term memory neural network (LSTM) model, used 3 million historical operation data of the distribution network as training corpus, and automatically generated the distribution network operation analysis report through sequence-to-sequence learning, with an average generation speed of 1,000 words per second. We built a Jenkins continuous integration environment, used the SonarQube code quality management platform, checked the automatically generated report source code for code specifications, security, reliability, etc., and found and fixed more than 30 code defects and security vulnerabilities.Using the deep learning library PyTorch, a Transformer text polishing model was built. With 5,000 manually written high-quality distribution network analysis reports as training materials, the automatically generated reports were corrected, semantically optimized, and language polished, which improved the readability of the reports by more than 25% on average. The generated HTML reports were converted to PDF format using the ApachePDFBox library, and the charts, formulas, tables, and other contents in the reports were identified and extracted using the OpenCV computer vision library. The report catalog and index were automatically generated to form an interactive and searchable electronic report format. On the Activiti workflow platform, based on the BPMN2.0 standard, the approval process for the distribution network redundancy analysis report was defined, including 6 links such as expert review, countersignature, and final review. The average approval time was shortened to less than 2 working days. The Kafka distributed message flow platform was used to realize real-time feedback and triggering of report approval opinions, and it was connected with the JIRA task management system to realize closed-loop tracking management of problems and defects.

[0048] The above contents are merely examples and explanations of the structure of the present invention. The technicians in this technical field may make various modifications or additions to the specific embodiments described or replace them in a similar manner. As long as they do not deviate from the structure of the invention or exceed the scope defined by the claims, they should all fall within the protection scope of the present invention.< / circle>

Claims

1. A method for automatically generating a feasibility study report for a distribution network based on an LDA probability model, characterized in that: The method comprises: Obtain the ledger information of power grid equipment and its electrical topology connection relationship, and generate a ledger document containing equipment name, type, location, and connection relationship information; The obtained ledger attributes and electrical topology relationships are used to construct a graph structure of the benchmark power grid, and the backup lines and backup equipment data are annotated to generate an annotation document containing the benchmark power grid structure and annotation information; According to the backup line and backup equipment data in the benchmark power grid structure, combined with the number of alternative paths between different equipment and lines, the redundancy of each node in the distribution network is calculated, and an electrical topology diagram containing redundancy information and a redundancy information document containing redundancy calculation results is generated; Based on the redundancy index calculated by the nodes, analyze the fault recovery capability of the distribution network at different redundancy levels, evaluate its switching and recovery speed when a fault occurs, and generate a fault analysis document containing the fault recovery capability analysis results and the switching and recovery speed evaluation results when a fault occurs; Based on the historical fault data of the distribution network and the redundancy index, the resource utilization of the distribution network is evaluated, the idleness of the backup equipment and lines in the non-fault state is determined, and an evaluation document containing the historical fault data and resource utilization evaluation results is generated; Summarize the ledger documents, annotation documents, redundancy information documents, fault analysis documents and evaluation documents of power grid equipment, perform preprocessing, build a vocabulary, represent each document as the word frequency vector of the word in the vocabulary, generate a document-word frequency matrix, the rows of the matrix represent documents, the columns represent words, and the values ​​in the matrix represent the frequency of the words in the document; Train the LDA model, set the number of topics, perform iterative training, and extract the topic distribution of each document and the vocabulary distribution of each topic; Based on the redundancy information, a feasibility report generation model is constructed. The feasibility report generation model is used to analyze the potential topics and patterns in the redundancy information and generate a distribution network redundancy feasibility report.

2. The method according to claim 1, wherein: The acquisition of grid equipment ledger information and its electrical topology connection relationship, and the generation of a ledger document containing equipment name, type, location, and connection relationship information, include: Acquire and analyze the model file, extract the device symbols and text annotation information in the model file, obtain the attribute information of the name, type and location of the power grid equipment, and construct the device attribute table; Analyze the graphic elements of the connection lines and arrows in the drawing file, combine the device location information in the device attribute table, determine the electrical connection relationship between the devices, and build a device topology connection relationship table; Integrate the equipment attribute table and equipment topology connection relationship table to form a ledger document; Perform attribute analysis on the generated ledger documents; Based on the attribute analysis results, the implicit relationships and frequent patterns between attributes are mined to find out the relevant attribute combinations that can form contextual logical connections and thought chains; Using the relevant attribute combinations found, combined with the business characteristics and professional knowledge of power grid equipment, the ledger documents are expanded and / or refined; Perform logic checks and consistency checks on the expanded and / or refined ledger documents.

3. The method according to claim 1, wherein: The obtained ledger attributes and electrical topology relationships are constructed into a graph structure of the benchmark power grid, and the backup lines and backup equipment data are marked to generate a marked document containing the benchmark power grid structure and marked information, including: According to the extracted ledger attributes and electrical topological relationships, nodes and relationships are created, with devices as nodes and topological connection relationships as edges, to build a graph structure of the benchmark power grid, including marking corresponding labels on backup lines and backup equipment to distinguish between main lines and backup lines, as well as main equipment and backup equipment. At the same time, for backup lines, the corresponding main line information is marked to form a backup relationship mapping; Analyze the key nodes and key paths in the power grid topology, evaluate the importance and impact range of the nodes, and generate network characteristic parameters of power supply paths and load distribution based on the topology and equipment attributes; For the equipment nodes in the graph structure, the ledger attribute information is associated, and for the standby equipment, the attributes of its commissioning conditions and switching methods are associated to form a standby equipment attribute portrait; Integrate the graph structure with the geographic information of the equipment, map the nodes onto the map according to the geographical location attributes of the equipment, and form a power grid geographic topology map to show the spatial distribution and geographical wiring relationship of the power grid; Based on the graph structure of the constructed benchmark power grid, a labeling document containing device nodes, topological connections and backup labeling information is generated. The labeling document includes visual attributes such as the color and shape of the nodes to distinguish different types of devices and connections.

4. The method according to claim 1, wherein: The redundancy of each node in the distribution network is calculated based on the backup line and backup equipment data in the benchmark power grid structure, combined with the number of alternative paths between different devices and lines, and a redundancy information document containing an electrical topology diagram of redundancy information and a redundancy calculation result is generated, including: Analyze the benchmark power grid structure, extract the connection relationship between nodes in the distribution network, and construct an adjacency matrix representing the network topology; Traversing all reachable nodes of the adjacency matrix and recording the path from the starting node to each reachable node, obtaining a set of candidate paths between all nodes in the distribution network, each path in the set of candidate paths representing a potential power supply solution; According to the attribute information of the backup lines and backup devices, the backup paths are screened and sorted, the paths containing the backup lines and backup devices are preferentially selected, the paths corresponding to the faulty devices and faulty lines are excluded, and the valid backup path set for each node is obtained; Combining the effective candidate path set and the benchmark power grid structure, calculating the redundancy index of each node, wherein the redundancy index represents the power supply reliability of the node; According to the calculation results of the redundancy index, the distribution network nodes are classified and sorted to generate a distribution network redundancy level classification table; The calculation results of the node redundancy index are associated with the geographical topology information of the distribution network, and nodes of different redundancy levels are marked with different colors or patterns to generate an electrical topology map that integrates the redundancy information to show the power supply reliability distribution of the distribution network; Summarize various data and results in the node redundancy analysis process, including the node adjacency matrix, the alternative path set, the valid alternative path set, the node redundancy index, the redundancy classification level and the distribution network redundancy classification table, and generate an electrical topology diagram containing redundancy information and a redundancy information document containing redundancy calculation results.

5. The method according to claim 1, wherein: The redundancy index calculated based on the node is used to analyze the fault recovery capability of the distribution network at different redundancy levels, evaluate its switching and recovery speed when a fault occurs, and generate a fault analysis document including the fault recovery capability analysis results and the switching and recovery speed evaluation results when a fault occurs, including: Based on the redundancy level classification table of the distribution network, a fault simulation analysis is performed for each redundancy level, including calculating the power supply recovery time of the node in the event of a fault according to the number of alternative paths and power supply capacity of the node, and obtaining the fault recovery time data under different redundancy levels; Analyze the impact of automatic switching mode and manual switching mode on fault switching time and obtain fault switching time data; Using the fault recovery time data and the fault switching time data, the fault recovery capability of the distribution network at different redundancy levels is evaluated, including the calculation of the fault recovery capability index; Analyze the correlation between the fault recovery capability index and the redundancy index, fit the functional relationship between the redundancy index and the fault recovery capability index, obtain the influence of the redundancy index on the fault recovery capability, and form a quantitative relationship model between the redundancy index and the fault recovery capability; The evaluation results of the fault recovery capability of the distribution network at different redundancy levels and the quantitative relationship model between the redundancy index and the fault recovery capability are summarized to generate a fault analysis document including the fault recovery capability analysis results and the evaluation results of the switching and recovery speed when a fault occurs.

6. The method according to claim 1, wherein: The method of evaluating the resource utilization of the distribution network based on the historical fault data of the distribution network in combination with the redundancy index, judging the idleness of the backup equipment and lines in a non-fault state, and generating an evaluation document containing the historical fault data and the resource utilization evaluation results includes: Obtaining historical fault records of the distribution network, wherein the fault records include fault occurrence time, faulty equipment, fault type, impact scope, and repair time information; Preprocessing the historical fault records of the distribution network; The distribution network fault data set is formed by summarizing the pre-processed historical fault records of the distribution network; According to the topological structure and equipment inventory information of the distribution network, the distribution network fault dataset is spatially mapped to determine the specific location of each fault event; Associating the specific location with the attribute information of the faulty device to form an attribute profile of the faulty device; Use the attribute portrait of the faulty equipment to analyze the fault characteristics and patterns of different types of equipment; Based on the redundancy information document, the distribution network fault data set is analyzed for correlation, and the failure rate, failure duration and failure impact range indicators at each redundancy level are calculated to evaluate the impact of the redundancy level on the reliability of the distribution network. Establish a quantitative relationship between redundancy and reliability based on the evaluation results; Analyze the relationship between faults, types and redundancy to find out the main causes and risk factors of faults; Evaluate the utilization efficiency of backup resources and identify the optimal balance between redundancy and resource utilization; Build a fault prediction model to predict the probability and type of equipment failure; Summarize the analysis results of the failure characteristics and laws of different types of equipment, the quantitative relationship between redundancy and reliability, the main causes and risk factors of failures, the evaluation results of backup resource utilization efficiency and failure prediction results, and generate an evaluation report on the distribution network's historical fault data analysis and resource utilization results.

7. The method according to claim 1, wherein: The method summarizes the ledger documents, annotation documents, redundancy information documents, fault analysis documents and evaluation documents of the power grid equipment, performs preprocessing, constructs a vocabulary, represents each document as a word frequency vector of the vocabulary in the vocabulary, generates a document-word frequency matrix, the rows of the matrix represent documents, the columns represent words, and the values ​​in the matrix represent the frequency of the words in the document, including: Summarize the ledger documents, annotation documents, redundancy information documents, fault analysis documents, and evaluation documents of power grid equipment, convert the summarized document contents into plain text format, and remove non-text elements, including pictures and tables, to obtain a plain text data set; Preprocessing the plain text data set, wherein the preprocessing includes word segmentation, stop word removal and part-of-speech tagging operations; Perform part-of-speech tagging on the preprocessed plain text dataset to form a vocabulary; Based on the word frequency statistics in the vocabulary, words with a frequency exceeding a threshold are selected as keywords to form a keyword list; By traversing each document, counting the frequency of each word in the keyword list in the document, a word frequency vector of the document is formed, wherein the dimension of the word frequency vector is equal to the length of the keyword list, and each element of the word frequency vector represents the frequency of the corresponding keyword in the document; A document-term frequency matrix is ​​constructed based on the term frequency vectors of each document. The rows of the matrix represent documents, the columns represent terms, and the values ​​in the matrix represent the frequency of the terms appearing in the document.

8. The method according to claim 1, wherein: The training of the LDA model, setting the number of topics, performing iterative training, and extracting the topic distribution of each document and the vocabulary distribution of each topic include: Based on the document-word frequency matrix, the input of the LDA topic model is constructed. Each document is represented as a frequency vector or weighted vector of the word and is normalized to ensure that the sum of the vector elements is 1. Combine document characteristics and domain knowledge to set the number of topics in the LDA model; Initializing parameters of the LDA model, including topic-word distribution parameters and document-topic distribution parameters; Train the LDA model and gradually converge to the optimal topic-word distribution and document-topic distribution; Calculate the posterior probability of each topic and update the topic-word distribution and document-topic distribution; Based on the updated topic-word distribution and document-topic distribution, the vocabulary distribution of each topic and the topic distribution of each document are obtained.

9. The method according to claim 1, wherein: Based on the redundancy information, a feasibility report generation model is constructed. The feasibility report generation model is used to analyze potential topics and patterns in the redundancy information and generate a distribution network redundancy feasibility report, including: Based on the redundancy information of each link in the distribution network, the LDA topic model is trained to obtain the indicator distribution of each topic and the topic distribution of each device or node, and to mine the potential semantic structure and association pattern in the redundancy information; Analyze the distribution of themes and indicators, explain and name each theme, summarize the redundancy patterns and characteristics represented by the themes, evaluate the rationality and practical significance of each theme, and screen out representative redundancy themes; Using the subject distribution of devices or nodes, devices or nodes with similar redundancy characteristics are classified into the same category to form a classification of redundant devices or nodes, and the redundancy level and risk degree of each category are analyzed; Based on the redundancy patterns mined by the LDA topic model and the equipment / node classification results, a redundancy feasibility report generation model is constructed to generate a corresponding feasibility analysis report based on the topic distribution characteristics of the equipment or node; It also includes: the feasibility report generation model combines with the local client or database to generate a distribution network redundancy feasibility report, the distribution network redundancy feasibility report includes the redundancy calculation results, resource utilization analysis and feasibility evaluation, and performs rule verification.

10. The method according to claim 9, wherein: The feasibility report generation model generates a distribution network redundancy feasibility report in combination with a local client or database. The distribution network redundancy feasibility report includes redundancy calculation results, resource utilization analysis and feasibility evaluation, and performs rule verification, including: Read data including equipment records, topology, and redundancy calculation results of the distribution network; Call the feasibility report generation model, fill in the predefined report template, and generate the first draft of the distribution network redundancy analysis report; The draft of the generated distribution network redundancy analysis report is associated and integrated with the distribution network resource utilization analysis results to evaluate the balance between the distribution network redundancy level and resource utilization efficiency, and form a feasibility analysis conclusion; Integrate the generated draft of the distribution network redundancy analysis report with the feasibility analysis conclusions; Conduct rule verification on the feasibility analysis report of redundancy of the integrated distribution network; Optimize and polish the feasibility analysis report of distribution network redundancy after rule verification; Digitally prepare the optimized and polished distribution network redundancy feasibility analysis report; The prepared feasibility analysis report on redundancy of distribution network shall be checked and approved to form the final feasibility analysis report on redundancy of distribution network.

Citation Information

Patent Citations

  • Power network power-cut scheme based major network N-1 risk assessment method and system

    CN102938106A

  • Method and device for analyzing self-healing capability of power distribution network based on data driving

    CN117955084A

  • Data-driven offline and online integrated simulation system and method for power distribution network

    WO2023115842A1

Cited By

  • Distribution network graph model data intelligent management system

    CN119809112A

  • A power distribution network diagram data intelligent management system

    CN119809112B

  • Portable data acquisition and analysis device for fault diagnosis of industrial equipment

    CN120029235A

  • Security report generation system and method, medium, electronic device and program product

    CN120185894A

  • System and method for converting electric power notification text into 5G message

    CN120429350A