A topology intelligent device management method and system based on a graph database

Abstract

The application discloses a kind of topological intelligent device management methods based on graph database, it is related to substation equipment management field, including collecting topological connection data;Topology connection data is parsed using graph database technology to establish substation equipment graph model;From topological analysis, the feature set of equipment is extracted, and the equipment is grouped and clustered using hierarchical clustering algorithm to find the implicit commonality and difference of equipment;Based on clustering result, key words or mode are identified from equipment name using intelligent classification algorithm, and the equipment is further classified to establish the definition and attribution of each equipment type;According to the equipment group and attribution result after intelligent classification, corresponding naming rule is formulated.The application stores the complex topological connection between equipment using graph database, can quickly analyze topological characteristics based on graph computing theory, and realizes the accurate judgment of node importance.

Description

Technical Field

[0001] This invention relates to the field of substation equipment management, and in particular to a topology intelligent equipment management method and system based on graph database. Background Technology

[0002] Against the backdrop of the digital transformation of power systems, the scale of equipment in substations is constantly expanding, and the types of equipment are becoming increasingly complex, bringing new challenges to the centralized and intelligent management of equipment. One major problem is the inconsistency in equipment naming conventions. Different manufacturers, regions, and commissioning times lead to inconsistent naming standards, resulting in difficulties in equipment data integration and management. Another problem is the limitation of traditional relational databases, which cannot handle large-scale equipment data and meet real-time requirements. Simultaneously, the lack of in-depth analysis and centralized management of equipment specifications hinders the intelligentization of equipment monitoring and maintenance. Furthermore, the lack of support for new smart grid technologies and a unified equipment information model standard also limits the security, scalability, and efficiency of information sharing in equipment management. Summary of the Invention

[0003] In view of the problems faced by substation equipment management in the context of the current digital transformation of the power system, such as differences in equipment naming rules, database limitations, and the lack of intelligent management, this invention is proposed.

[0004] Therefore, the problem to be solved by this invention is how to achieve effective management and intelligent processing of device topology connection data through methods such as topology analysis, intelligent classification and naming rule formulation.

[0005] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0006] In a first aspect, embodiments of the present invention provide a topology intelligent device management method based on a graph database, which includes collecting topology connection data of a 220kV substation in a target area and inputting it into a graph database for parsing and preprocessing; using graph database technology to parse the topology connection data to establish a substation equipment graph model and performing topology analysis to identify the connection relationships and equipment groups between devices; extracting the feature set of devices from the topology analysis and using a hierarchical clustering algorithm to group and cluster the devices to discover the implicit commonalities and differences of the devices; based on the clustering results, applying an intelligent classification algorithm to identify keywords or patterns from the device names, further classifying the devices to establish the definition and affiliation of each device type; formulating corresponding naming rules according to the intelligently classified device groups and affiliation results, and using the naming rules to name the devices in each device group.

[0007] As a preferred embodiment of the graph database-based topology intelligent device management method of the present invention, the method includes the following steps: performing topology analysis to identify the connection relationships between devices and device groups: importing preprocessed topology connection data into a graph database and constructing device nodes according to the device attributes; establishing a substation device graph model based on the topology connection data to represent the connection relationships between device nodes, and adding edges to represent the connection relationships between devices; performing topology analysis on the established device graph model to identify the connection relationships between devices; and performing device group analysis, using algorithms and indicators to calculate the importance and correlation of device nodes to identify device groups and key devices.

[0008] As a preferred embodiment of the graph database-based topology intelligent device management method of the present invention, the method for calculating the importance and correlation of device nodes using algorithms and indicators to identify device groups and key devices includes the following steps: querying the direct connection relationships between device nodes by traversing the edges in the graph model; finding paths between devices based on an iteratively deepening depth-first search algorithm to identify indirect connection relationships between devices; constructing an adjacency matrix A of the device graph and calculating the degree centrality of the nodes; for each node n i Iterate through all its neighbor nodes n j Calculate the degree of each neighboring node and sum them to obtain node n. i The proximity centrality is calculated; the shortest path algorithm is used to calculate the shortest path between any two nodes, the number of shortest paths for each pair of nodes is counted, and the shortest path numbers for all pairs of nodes are summed to calculate the betweenness centrality of the nodes; the eigenvector centrality is calculated by calculating the maximum eigenvalue and the corresponding eigenvector of the adjacency matrix A; and the key equipment is determined based on the degree centrality, proximity centrality, betweenness centrality, and eigenvector centrality of the nodes.

[0009] As a preferred embodiment of the graph database-based topology intelligent device management method of the present invention, the method for finding paths between devices and identifying indirect connections between devices based on an iterative deepening depth-first search algorithm includes the following steps: determining the starting device node and the target device node for finding the path, and setting the initial iteration level to 1; starting from the starting device node, performing a depth-first search within the current iteration level; starting from the current node, traversing all neighboring nodes in the current level; for each neighboring node, determining whether it is the target node, and if it is, recording the path; and determining whether the current iteration level has been reached. The maximum number of nodes is determined by the algorithm. If all nodes in the current level have been traversed but no path is found, the iteration level is increased by 1, and the starting device node is returned to proceed to the next level. For unvisited neighbor nodes, the above steps are recursively executed, and the depth-first search continues. If a path is found in the current iteration level, the search stops, the previous level is backtracked, and other paths are searched. The above steps are repeated, increasing the iteration level, until the maximum level limit is reached or the path to the target node is found. When the maximum level limit is reached or the path to the target node is found, the found path result is returned; if no path is found, an empty result is returned.

[0010] As a preferred embodiment of the graph database-based topology intelligent device management method of the present invention, the specific formula for the degree centrality of a node is as follows:

[0011]

[0012] Among them, V i Represents a node, k i The degree of a node is represented by N, the number of nodes is represented by N-1, and the maximum degree value exists.

[0013] node n i The specific formula for proximity centrality is as follows:

[0014]

[0015] Where, n i Let i represent node i, N represent the number of nodes, j represent neighboring nodes, and d represent the number of neighboring nodes. ij This represents the distance from node i to its neighboring node j.

[0016] As a preferred embodiment of the graph database-based topology intelligent device management method of the present invention, the method for calculating the shortest path between any two nodes using the shortest path algorithm includes the following steps: initializing a two-dimensional array to represent the shortest path distance between any two nodes; if there is an edge from node i to node j, the distance is set as the edge weight; if there is no edge from node i to node j, the distance is set to infinity; for each node k, using this node as a transit point, traversing all node pairs (i,j); if the path through node k is shorter than the current shortest path distance, updating the shortest path distance array; after traversing all nodes, the final shortest path distance array will store the shortest path length between any two nodes; for the shortest path between each pair of nodes, counting the number of nodes traversed and recording it; traversing the shortest paths of all node pairs, summing the number of shortest paths between each pair of nodes to obtain the betweenness centrality of the nodes.

[0017] As a preferred embodiment of the graph database-based topology intelligent device management method of the present invention, the method for identifying keywords or patterns from device names using an intelligent classification algorithm includes the following steps: analyzing device name keywords and comparing them with keywords of known device types; if a device name contains keywords corresponding to a certain device type, the device type to which the device belongs is preliminarily determined; if the keywords in the device name are ambiguous or misleading compared with the keywords of known device types, the device is further analyzed in conjunction with device attribute information, and the device attributes are compared with the characteristics of known device types; if the device attributes match the characteristics of a certain device type, the device type to which the device belongs is preliminarily determined; if the device attributes do not completely match the characteristics of known device types, the connection relationship between the device and other devices and the device's position in the topology connection graph are considered; a comprehensive judgment is made by combining domain expertise, experience, and actual situation to determine the final device type; based on the intelligent classification results, the devices are grouped into different categories or types, and each device group should contain devices with similar characteristics or belonging to the same category; within each device group, corresponding naming rules are formulated to ensure the consistency and understandability of device naming.

[0018] Secondly, embodiments of the present invention provide a topology intelligent device management system based on a graph database, comprising: a data model module for collecting topology connection data of a 220kV substation in a target area and inputting it into a graph database for parsing and preprocessing; a graph model construction module for using graph database technology to parse the topology connection data to establish a graph model of substation equipment and performing topology analysis to identify the connection relationships and equipment groups between devices; a clustering analysis module for extracting feature sets of devices from the topology analysis and using hierarchical clustering algorithms to group and cluster the devices to discover the implicit commonalities and differences of the devices; an equipment type attribution module for using intelligent classification algorithms to identify keywords or patterns from the device names based on the clustering results, further classifying the devices to establish the definition and attribution of each equipment type; and a naming confirmation module for formulating corresponding naming rules based on the intelligently classified device groups and attribution results.

[0019] Thirdly, embodiments of the present invention provide a computer device, including a memory and a processor, wherein the memory stores a computer program, wherein: when the computer program instructions are executed by the processor, they implement the steps of the graph database-based topology intelligent device management method as described in the first aspect of the present invention.

[0020] Fourthly, embodiments of the present invention provide a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program instructions are executed by a processor, they implement the steps of the graph database-based topology intelligent device management method as described in the first aspect of the present invention.

[0021] The beneficial effects of this invention are as follows: This invention uses a graph database to store complex topological connections between devices, and based on graph computing theory, it can quickly analyze topological features and accurately determine the importance of nodes; it uses machine learning to automatically summarize naming patterns from massive amounts of device data, overcoming the problem of non-standard naming caused by regional and manufacturer differences; it significantly improves the level of precision and intelligence in equipment monitoring, control, and maintenance within substations, effectively supporting power grid companies in implementing precise operation and maintenance, ensuring the safe and efficient operation of the power grid, and promoting the digital transformation of the power system. Attached Figure Description

[0022] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Wherein:

[0023] Figure 1 This is a flowchart of a topology-based intelligent device management method based on graph databases.

[0024] Figure 2 This is a computer device graph for a topology-based intelligent device management method using graph databases. Detailed Implementation

[0025] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0026] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0027] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0028] Example 1

[0029] Reference Figures 1-2 This is the first embodiment of the present invention, which provides a topology intelligent device management method based on a graph database, including:

[0030] S1: Collect the topology connection data of the 220kV substations in the target area and input it into the graph database for parsing and preprocessing.

[0031] Specifically, the system collects and parses massive amounts of 220kV substation topology connection data from a certain region into a graph database, and then imports the structured files of all 220kV substations in the region into a summary database. The format is not limited to EXCEL, XML, CIM / E, JSON, etc., and the substations include regional affiliation information.

[0032] It should be noted that data access can be achieved using the following three typical methods, but is not limited to other access methods. 1. Manual Input: When the required data source format cannot be automatically recognized by the software, data can be entered item by item through the manual interface or input screen. The system will verify the input data and automatically load the data into the data model module after successful verification. 2. Data File Import: When the required data source is a structured file format (such as EXCEL, XML, CIM / E, JSON, etc.), the file can be imported through the software's file loading interface. This module will automatically identify and interpret the data format based on the file type, perform data verification, and automatically load the data into the data model module after successful verification. 3. Third-Party Software System Interface Input: When the required data source comes from a third-party software system, it can be obtained through database access interfaces, communication protocols, API interfaces, etc. This module will automatically identify and interpret the data format based on the obtained data, perform data verification, and automatically load the data into the data model module.

[0033] S2: Use graph database technology to parse topology connection data to build a graph model of substation equipment, and perform topology analysis to identify the connection relationships between equipment and equipment groups.

[0034] Specifically, it includes the following steps:

[0035] S2.1: Import the preprocessed topology connection data into the graph database and construct device nodes according to the device attributes.

[0036] Specifically, first, determine the graph database software to be used and install and configure it accordingly; then, create device nodes according to the requirements of device attributes and associate device attributes with nodes; finally, create the connection relationships between device nodes based on the imported topology connection data, such as physical connections, control connections, and operation and maintenance connections.

[0037] S2.2: Based on the topology connection data, establish a substation equipment graph model to represent the connection relationship between equipment nodes, and add edges to represent the connection relationship between equipment.

[0038] Specifically, the model file of the 220kV substation is analyzed. Based on the established equipment nodes and connection relationships, a substation equipment model is established in the graph database. In the graph database, node switch diagram models of equipment such as busbars, switches and disconnectors are established respectively.

[0039] S2.3: Perform topology analysis on the established device diagram model to identify the connection relationships between devices.

[0040] S2.4: Perform equipment group analysis, using algorithms and metrics to calculate the importance and correlation of equipment nodes to identify equipment groups and key equipment.

[0041] Specifically, it includes the following steps:

[0042] S2.4.1: Query the direct connection relationships between device nodes by traversing the edges in the graph model.

[0043] Specifically, the query function of the graph database is used to determine the direct connection relationships between device nodes, including shared edges and associated attributes.

[0044] S2.4.2: A depth-first search algorithm based on iterative deepening is used to find paths between devices to identify indirect connections between them.

[0045] Specifically, determine the starting and target device nodes for the path to be found, and set the initial iteration level to 1; starting from the starting device node, perform a depth-first search within the current iteration level; starting from the current node, traverse all neighboring nodes in the current level; for each neighboring node, determine if it is the target node, and if so, record the path; determine if the maximum number of nodes in the current iteration level has been reached, and if all nodes in the current level have been traversed but no path has been found, increment the iteration level by 1 and return to the starting device node to enter the next level; for unvisited neighboring nodes, recursively execute the above steps and continue the depth-first search; if a path is found within the current iteration level, stop the search and backtrack to the previous level to continue searching for other paths; repeat the above steps, increasing the iteration level, until the maximum level limit is reached or the path to the target node is found; when the maximum level limit is reached or the path to the target node is found, return the found path result; if no path is found, return an empty result.

[0046] S2.4.3: Identify device groups and key devices by calculating the importance and relevance of device nodes based on algorithms and metrics.

[0047] Specifically, the process includes the following steps:

[0048] S2.4.3.1: Construct the adjacency matrix A of the device graph and calculate the degree centrality of the nodes.

[0049] Specifically, when constructing the adjacency matrix A of the device graph, nodes are represented as rows and columns of the matrix, A ij A represents the connection weight between nodes i and j. If there is a connection between nodes i and j, then A... ij Let A be the weight value for the connection. If there is no connection between nodes i and j, then A... ij It is 0.

[0050] Furthermore, we traverse all nodes and count the degree of each node (that is, the number of edges directly connected to the node). A node with higher degree centrality means that it has more connections to other nodes and therefore has greater influence in the network.

[0051] Specifically, the formula for degree centrality is as follows:

[0052]

[0053] Where, n i Representing nodes i and k i The degree of a node is represented by N, the number of nodes is represented by N-1, and the maximum degree value exists.

[0054] S2.4.3.2: For each node n i Iterate through all its neighbor nodes n j Calculate the degree of each neighboring node and sum them to obtain node n. i proximity centrality.

[0055] Preferably, the specific formula for the proximity centrality of a node is as follows:

[0056]

[0057] Where, n i Let i represent node i, N represent the number of nodes, j represent neighboring nodes, and d represent the number of neighboring nodes. ij This represents the distance from node i to its neighboring node j.

[0058] It should be noted that proximity centrality measures how closely a node is connected to its neighboring nodes; a higher value indicates a stronger connection with other nodes.

[0059] S2.4.3.3: Use the shortest path algorithm to calculate the shortest path between any two nodes, count the number of shortest paths for each pair of nodes, and sum the shortest path counts for all pairs of nodes to calculate the betweenness centrality of the nodes.

[0060] Preferably, a two-dimensional array is initialized to represent the shortest path distance between any two nodes. If there is an edge between node i and node j, the distance is set as the edge weight; if there is no edge between node i and node j, the distance is set to infinity. For each node k, this node is used as a transit point, and all node pairs (i,j) are traversed. If the path through node k is shorter than the current shortest path distance, the shortest path distance array is updated. After traversing all nodes, the final shortest path distance array will store the shortest path length between any two nodes. For the shortest path between each pair of nodes, the number of nodes traversed (excluding the starting node and the target node) is counted and recorded. The shortest paths between all node pairs are traversed, and the number of shortest paths between each pair of nodes is added together to obtain the betweenness centrality of the nodes.

[0061] S2.4.3.4: Calculate the eigenvector centrality by calculating the largest eigenvalue and the corresponding eigenvector of the adjacency matrix A.

[0062] Specifically, the elements in the eigenvector x represent the eigenvector centrality of each node. The larger the eigenvalue, the higher the centrality of the node.

[0063] S2.4.3.5: Determine key equipment based on the degree centrality, proximity centrality, betweenness centrality, and eigenvector centrality of nodes.

[0064] Furthermore, key devices are identified based on node centrality metrics. Nodes with high degree centrality indicate more direct connections to other nodes and are likely key devices; nodes with high proximity centrality indicate close connections to their neighbors and are also likely key devices; nodes with high betweenness centrality indicate that they act as bridges connecting other nodes in the network and are also likely key devices; nodes with high eigenvector centrality may play an important role in the global network structure and are also likely key devices.

[0065] S3: Extract the feature set of devices from topology analysis, and use hierarchical clustering algorithm to group and cluster the devices to discover the implicit commonalities and differences of the devices.

[0066] Specifically, node features, connection features, and attribute features are extracted from the equipment graph model based on topology analysis. These features are then integrated into a unified feature representation of the equipment. The attribute features include voltage level, equipment type, city / region, equipment manufacturer, and year of commissioning.

[0067] Specifically, clustering using a hierarchical clustering algorithm includes the following steps: calculating the cosine similarity between device feature vectors to construct a similarity matrix; dividing device names and numbers into different clusters based on the selected features and initializing the clusters; determining the two clusters with the highest similarity based on the similarity matrix and merging them into a new cluster; updating the similarity matrix and adjusting the similarity between the merged cluster and other clusters; repeating the steps until all devices are merged into one cluster or a pre-set upper limit for the number of clusters is reached to obtain clustering results at different levels; and analyzing the clustering results to interpret the meaning of the device names and numbers in each cluster.

[0068] S4: Based on the clustering results, apply intelligent classification algorithms to identify keywords or patterns from the device names, and further classify the devices to establish the definition and affiliation of each device type.

[0069] Specifically, it includes the following steps:

[0070] S4.1: Collect the names of all devices, perform word segmentation, and train word vectors using a word vector model.

[0071] Specifically, the names of all devices are collected as text data, and each device name is segmented (using a word segmentation tool); each word is converted into a corresponding word vector using a pre-trained word vector model; for device names, the word vector representation of the name is obtained by taking a weighted average of the word vectors, and this is used as the feature vector of the device name.

[0072] S4.2: Based on name word vectors, use the KNN algorithm to classify devices.

[0073] Specifically, a sample set is constructed containing some device names and their corresponding manually labeled device types, with each sample having a device name and a device type label; the word vector representation of each name is calculated based on the word vectors of the device names obtained in S4.1; for a new device, the device name is converted into a word vector, and the similarity is calculated with the word vectors of all names in the sample set to obtain a similarity vector; the K samples in the sample set with the highest similarity to the new device name are selected, and the Euclidean distance in the KNN algorithm is used as the similarity measure; the type of the new device is inferred from the device type labels in these K most similar samples through majority voting.

[0074] S4.3: Analyze the correspondence between equipment name keywords and types, and adjust the equipment type classification based on the name keywords.

[0075] Specifically, for each equipment type, the most frequent keywords in the names of equipment of that type in the statistical sample set are used as characteristic words representing that type. The equipment type is defined based on the correspondence between these keywords and equipment types. For equipment with unclear categories or incorrect classifications, the classification is corrected based on the correspondence between the keywords contained in its equipment name and the keywords obtained in S4.3, so as to adjust the classification of these equipment types to match the keywords in the name.

[0076] It should be noted that name keywords cannot fully represent the type of equipment. Some equipment names may be associated with multiple types, and there may be misleading or similar names. Therefore, when adjusting the equipment type classification based on name keywords, it is necessary to verify and confirm it by taking into account other information.

[0077] Preferably, the device name keywords are analyzed and compared with the keywords of known device types. If the device name contains keywords corresponding to a certain device type, the device type to which the device belongs is initially determined. If the keywords in the device name are ambiguous, vague, or misleading compared with the keywords of known device types, the device is further analyzed in conjunction with the device attribute information, and the device attributes are compared with the characteristics of known device types. If the device attributes match the characteristics of a certain device type, the device type to which the device belongs is initially determined. If the device attributes do not completely match the characteristics of known device types, the connection relationship between the device and other devices and the device's position in the topology connection diagram are considered. Devices of a certain type are usually closely connected to a specific device type and share a specific position. A comprehensive judgment is made by combining domain expertise, experience, and actual situation to determine the final device type.

[0078] S5: Develop corresponding naming rules based on the intelligently classified device groups and their affiliation results, and use the naming rules to name devices in each device group.

[0079] Specifically, based on the intelligent classification results, devices are grouped into different categories or types. Each device group should contain devices with similar characteristics or belonging to the same category. Within each device group, corresponding naming rules are established to ensure the consistency and understandability of device naming. For devices in each device group, they are named according to the naming rules.

[0080] It should be noted that, in this embodiment, the naming rule adopts a hierarchical naming structure, including device type abbreviation, device importance, device physical location, and device number, etc.

[0081] Furthermore, this embodiment also provides a topology intelligent device management system based on a graph database, including a data model module for collecting topology connection data of 220kV substations in a target area and inputting it into a graph database for parsing and preprocessing; a graph model construction module for using graph database technology to parse the topology connection data to establish a graph model of substation equipment and perform topology analysis to identify the connection relationships and equipment groups between equipment; a clustering analysis module for extracting the feature set of equipment from the topology analysis and using hierarchical clustering algorithms to group and cluster the equipment to discover the implicit commonalities and differences of the equipment; an equipment type attribution module for using intelligent classification algorithms to identify keywords or patterns from equipment names based on the clustering results, further classifying the equipment to establish the definition and attribution of each equipment type; and a naming confirmation module for formulating corresponding naming rules based on the intelligently classified equipment groups and attribution results.

[0082] This embodiment also provides a computer device applicable to the topology intelligent device management method based on graph database, including a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the topology intelligent device management method based on graph database as proposed in the above embodiment.

[0083] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0084] This embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements the graph database-based topology intelligent device management method as proposed in the above embodiments.

[0085] In summary, this invention employs a graph database to store complex topological connections between devices. Based on graph computing theory, it can quickly analyze topological features and accurately determine the importance of nodes. It uses machine learning to automatically summarize naming patterns from massive amounts of device data, overcoming the naming inconsistencies caused by regional and manufacturer differences. It significantly improves the precision and intelligence of equipment monitoring, control, and maintenance within substations, effectively supporting power grid companies in implementing precise operation and maintenance, ensuring the safe and efficient operation of the power grid, and promoting the digital transformation of the power system.

[0086] Example 2

[0087] Reference Figures 1-2 This is the second embodiment of the present invention, which provides a topological intelligent device management method based on graph database. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through economic benefit calculation and simulation experiment.

[0088] Specifically, this invention takes a 220kV substation in a certain area as an example, collects the topology connection data of the 220kV substation in the target area, and imports it into a graph database for parsing via data file import. However, since the data may contain noise, missing values, or outliers, data cleaning and preprocessing are required, including handling missing values, deleting duplicate data, and standardizing data.

[0089] Furthermore, necessary nodes include substations, substation busbars, circuit breakers, disconnectors, etc., and each requires parameters such as substation manufacturer, commissioning date, equipment ID, and topology connection nodes. Other nodes and their attributes depend on specific requirements. Only two tables are shown here; Table 1 shows some substation data, and Table 2 shows some substation busbar data.

[0090] Table 1. Partial Substation Data

[0091] logo Original Chinese name ... highest voltage level Plant type The area identifier ... Commissioning date B1 Substation A ... 220kV substation City 1 ... 2015-01-01 B2 B Substation ... 220kV substation 2 prefecture-level cities ... 2016-03-21 B3 C Substation ... 220kV substation 3 prefecture-level cities ... 2016-09-25 B4 D Substation ... 220kV substation 4 prefecture-level cities ... 2017-02-14 B5 E Substation ... 220kV substation 5 prefecture-level cities ... 2017-07-11 ... ... ... ... ... ... ... ...

[0092] Table 2: Partial Substation Data

[0093] Attribute Name Device Unique Identifier Collection time Operational status ... Station ID Corresponding connection points <![CDATA[U a Current MB001 2023-02-15 10:00:00 In use ... B1 CP001 <![CDATA[U b Current MB002 2023-02-15 10:05:00 In use ... B1 CP002 <![CDATA[U c Current MB003 2023-02-15 10:10:00 In use ... B1 CP003 <![CDATA[I a Current MB101 2023-02-15 10:15:00 In use ... B2 CP101 <![CDATA[I b Current MB102 2023-02-15 10:20:00 In use ... B2 CP102 ... ... ... ... ... ... ...

[0094] Furthermore, based on the imported topology connection data, substation equipment nodes and their connection relationships are created in the Tigergraph graph database, and a preliminary equipment topology graph model is constructed. Taking substation B1 as an example, substation node B1, bus nodes MB001, MB002, etc. are created, and "connection" relationship edges are added to represent the physical connection between the two devices.

[0095] Furthermore, based on the data clustering results, the 220kV circuit breaker in City 1 and the adjacent disconnectors on both sides contain the same "three-digit number" in their names. The transformer switch, line switch, bus tie switch and sectionalizing switch are all named "three-digit number + switch". The switch numbers do not have obvious characteristics, and semantic recognition is needed to further distinguish whether they are transformer switches, line switches, bus tie switches or sectionalizing switches from their names.

[0096] Furthermore, based on the clustering results, the influencing factors of equipment naming and numbering rules are determined, and unified equipment naming and numbering rules are formulated. For example, similar equipment names and numbers of 220kV bus tie switches managed by the same municipal company can be grouped into the same category, and corresponding naming and numbering rules can be formulated for this category.

[0097] Preferably, the comparison indicators between the method of the present invention and the traditional method are shown in Table 3.

[0098] Table 3 Comparison Indicators between the Method of the Present Invention and Traditional Methods

[0099] Comparison indicators Our invention Traditional methods Equipment data query time 0.1s / time 2 seconds / time Equipment naming standardization rate 92％ 60％ Device relationship resolution success rate 90％ 70％ Key equipment identification accuracy 91％ 85％ Equipment classification accuracy 93％ 75％ Model training time Reduced by 30% No significant changes Naming rules Manual extraction automation

[0100] Furthermore, as can be seen from the comparison table, compared with traditional methods, this invention shows significant advantages in all aspects of equipment data management and intelligent analysis: the equipment data query time is shortened by nearly 20 times, from 2 seconds to 0.1 seconds each time, greatly improving query response efficiency; the equipment naming standardization rate is increased by 32 percentage points, from 60% to 92%, greatly improving the degree of automation; the success rate of equipment relationship resolution is increased to 90%, a significant improvement compared to the 70% of traditional methods; the accuracy of key equipment identification is improved by 6 percentage points, reaching 91%; the correctness of equipment classification is also increased by nearly 20 percentage points, achieving a high degree of accurate classification; the model training time is significantly shortened, improving efficiency compared to traditional methods; and the naming rule formulation is automated, while traditional methods rely on manual extraction.

[0101] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A topological intelligent device management method based on graph database, characterized in that: include, Collect the topology connection data of 220kV substations in the target area and input it into the graph database for parsing and preprocessing; Graph database technology is used to parse topology connection data to build a graph model of substation equipment, and topology analysis is performed to identify the connection relationships between equipment and equipment groups; The feature set of the devices is extracted from the topology analysis, and the devices are grouped and clustered using a hierarchical clustering algorithm to discover the implicit commonalities and differences among the devices. Based on the clustering results, an intelligent classification algorithm is applied to identify keywords or patterns from the device names, and the devices are further classified to establish the definition and classification of each device type. Based on the intelligently categorized device groups and their affiliation results, corresponding naming rules are formulated, and the naming rules are used to name devices in each device group; The step of performing topology analysis to identify connectivity relationships and device groups includes the following steps: Import the preprocessed topology connection data into the graph database and construct device nodes based on the device attributes; A substation equipment graph model is built based on topology connection data to represent the connection relationship between equipment nodes, and edges are added to represent the connection relationship between equipment. Perform topology analysis on the established device diagram model to identify the connection relationships between devices; Perform equipment cluster analysis, using algorithms and metrics to calculate the importance and correlation of equipment nodes in order to identify equipment clusters and key equipment; The process of using algorithms and metrics to calculate the importance and correlation of device nodes in order to identify device groups and key devices includes the following steps: Query the direct connection relationships between device nodes by traversing the edges in the graph model; A depth-first search algorithm with iterative deepening is used to find paths between devices to identify indirect connections between them; Construct the adjacency matrix of the device graph And calculate the degree centrality of the nodes; For each node traverse all its neighbor nodes Calculate the degree of each neighboring node and sum them to obtain the node. proximity centrality; The shortest path algorithm is used to calculate the shortest path between any two nodes, the number of shortest paths for each pair of nodes is counted, and the shortest path counts for all pairs of nodes are summed to calculate the betweenness centrality of the nodes. By calculating the adjacency matrix Given the largest eigenvalue and its corresponding eigenvector, calculate the eigenvector centrality. Key equipment is determined based on the degree centrality, proximity centrality, betweenness centrality, and eigenvector centrality of nodes; The depth-first search algorithm based on iterative deepening to find paths between devices and identify indirect connections between devices includes the following steps: Determine the starting and target device nodes for the path to be found, and set the initial iteration level to [value]. ; Starting from the initial device node, perform a depth-first search within the current iteration level; Starting from the current node, traverse all neighboring nodes at the current level; For each neighboring node, determine whether it is the target node; if it is the target node, record the path. Check if the maximum number of nodes in the current iteration level has been reached. If all nodes in the current level have been traversed but no path has been found, increase the iteration level. And return to the starting device node to enter the next level; For unvisited neighbor nodes, recursively execute the above steps and continue the depth-first search; If a path is found within the current iteration level, stop searching deeper, backtrack to the previous level, and continue searching for other paths; Repeat the above steps, increasing the iteration level, until the maximum level limit is reached or the path to the target node is found; When the maximum level limit is reached or the path to the target node is found, the found path result is returned; otherwise, an empty result is returned.

2. The topology intelligent device management method based on graph database as described in claim 1, characterized in that: The specific formula for the degree centrality of the node is as follows: ； in, Represents a node. Represents the degree of a node. Indicates the number of nodes. Indicates the maximum degree value that exists; The node The specific formula for proximity centrality is as follows: ； in, Representing node i, Indicates the number of nodes. Indicates neighboring nodes, Represents a node to neighboring nodes The distance.

3. The topology intelligent device management method based on graph database as described in claim 2, characterized in that: The process of calculating the shortest path between any two nodes using the shortest path algorithm includes the following steps: Initialize a two-dimensional array to represent the shortest path distance between any two nodes. If the nodes To the node If there is an edge, set the distance as the edge weight; if the node... To the node If there is no boundary, then set the distance to infinity; For each node Using this node as a transit point, traverse all node pairs. ; If it passes through the node If a path can be shorter than the current shortest path distance, then update the shortest path distance array; After traversing all the nodes, the final shortest path distance array will store the shortest path length between any two nodes; For the shortest path between each pair of nodes, count the number of nodes traversed and record it; Traverse the shortest paths between all node pairs and sum the number of shortest paths between each pair of nodes to obtain the betweenness centrality of the nodes.

4. The topology intelligent device management method based on graph database as described in claim 1, characterized in that: The application of intelligent classification algorithm to identify keywords or patterns from device names includes the following steps: Analyze the keywords in the equipment name and compare them with the keywords of known equipment types. If the equipment name contains keywords corresponding to a certain equipment type, then the equipment type to which this equipment belongs can be preliminarily determined. If the keywords in the device name are ambiguous or misleading compared to the keywords of the known device type, the device will be further analyzed in conjunction with the device attribute information, and the attributes of the device will be compared with the characteristics of the known device type. If the device attributes match the characteristics of a certain device type, then the device type to which this device belongs can be preliminarily determined; If the device attributes do not completely match the characteristics of the known device type, then consider the connection relationship between the device and other devices, as well as the device's position in the topology connection diagram; The final equipment type is determined by combining domain expertise, experience, and practical considerations. Based on the intelligent classification results, the devices are grouped into different categories or types. Each device group should contain devices with similar characteristics or belonging to the same category. Within each device group, establish corresponding naming rules to ensure consistency and understandability of device names.

5. A graph database-based topology intelligent device management system, based on the graph database-based topology intelligent device management method according to any one of claims 1 to 4, characterized in that: It also includes, The data model module is used to collect topology connection data of 220kV substations in the target area and input it into the graph database for parsing and preprocessing. The graph model building module is used to parse topology connection data using graph database technology to build a graph model of substation equipment, and to perform topology analysis to identify the connection relationships between equipment and equipment groups; The clustering analysis module is used to extract the feature set of devices from topology analysis and use hierarchical clustering algorithms to group and cluster the devices to discover the implicit commonalities and differences among them. The equipment type attribution module is used to identify keywords or patterns from equipment names based on clustering results and apply intelligent classification algorithms to further classify the equipment in order to establish the definition and attribution of each equipment type. The naming confirmation module is used to formulate corresponding naming rules based on the intelligently classified device groups and their affiliation results.

6. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, it implements the steps of the topology intelligent device management method based on graph database as described in any one of claims 1 to 4.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the steps of the topology intelligent device management method based on graph database as described in any one of claims 1 to 4.

Images