Distribution network dispatching knowledge fusion and graph storage method and system based on graph database

Abstract

The distribution network dispatching knowledge fusion and graph storage method and system based on the graph database abstracts the ontology and attributes in the distribution network dispatching entity knowledge graph into points and edges in the distribution network dispatching entity knowledge graph, and establishes a subgraph model of the distribution network dispatching entity knowledge graph; abstracts the events, relationship between things and other information in the distribution network dispatching knowledge graph into points and edges in the distribution network dispatching knowledge graph, and establishes a subgraph model of the distribution network dispatching knowledge graph; uses the knowledge fusion method to form a link path from the subgraph model of the distribution network dispatching knowledge graph to the subgraph model of the distribution network dispatching entity knowledge graph; stores the distribution network dispatching entity knowledge graph and the distribution network dispatching knowledge graph after knowledge fusion in the graph database according to the field type. The present invention solves the technical bottlenecks of the existing distribution network dispatching knowledge system in the application of systematic knowledge linking, auxiliary decision-making reference query, etc.

Description

Technical Field

[0001] The present invention relates to the field of smart grid dispatching technology, and more specifically, to a distribution network dispatching knowledge fusion and graph storage method and system based on a graph database. Background Art

[0002] Grid dispatching has high standards for efficient and orderly emergency response. With the rapid development of the power system, the structure and operation mode of the distribution network have become more complex. The network analysis after the fault has problems such as unclear topology and inaccurate data. The difficulty of fault handling is increasing. It is difficult to accurately warn, classify and infer faults and handle them by relying on traditional grid analysis methods. The experience or handling decision-making plans accumulated by the dispatcher can be formed into an orderly structured knowledge base and can be stored, queried and inferred, which is of great significance to the fault decision-making of distribution network dispatching.

[0003] The knowledge graph technology is used to store the power grid equipment, structure, and fault information, and the knowledge graph is used to extract, represent, and manage knowledge to assist dispatchers in handling faults, which can effectively improve the power grid's emergency handling capabilities and the level of intelligent dispatching.

[0004] Prior Art 1 (CN107330125B) "Massive unstructured distribution network data integration method based on knowledge graph technology" collects unstructured distribution network data of each information system by a data acquisition unit, and performs quality analysis and data cleaning on the unstructured distribution network data of each information system respectively; constructs a local data index based on the local knowledge graph according to the processed unstructured distribution network data of each information system; sends the local data index based on the local knowledge graph to the data management center through a big data connector; the data management center constructs a global data index based on the global knowledge graph. Prior Art 1 unifies the local index information of each entity in the local data index table according to the entities, classes, attribute values ​​and hierarchical relationships of the local knowledge graph obtained during the detection and elimination of conflicts, and constructs a global knowledge graph; constructs a mapping relationship between the global knowledge graph and the local knowledge graphs of each information system. The mapping relationship constructed in Prior Art 1 still exists between entity graphs and does not involve the matter graph. Prior art 2 (CN113868429A) "A method and system for constructing and integrating security knowledge graphs based on geographic knowledge" obtains geographic data and security data, stores the two types of data as ontology in a graph database; analyzes the correlation between the security scenario ontology and the geographic location ontology; adds and integrates the geographic location ontology with a greater correlation with the security scenario ontology into the knowledge graph node of the security ontology; redistributes the orientation relationship, distance relationship, inheritance relationship, and affiliation relationship of the knowledge graph according to whether the subsequent use results of the security knowledge graph ontology meet the security requirements; highlights the associated geographic ontology nodes in the security ontology according to the probability of danger during the knowledge graph display process, so that managers can obtain the most dangerous geographic location information in a timely manner in an emergency state of a security incident. Prior art 2 realizes the fusion of geographic knowledge graphs and security knowledge graphs. The graph fusion scheme of prior art 2 only considers the matching of entities, but there is still a problem that different names refer to the same entity between actual graphs, so prior art 2 has the phenomenon of missing or wrong matching.

[0005] Relying on a large amount of manpower to extract triples, the construction cost of the knowledge graph is very high. The solution that relies on manual judgment of the relationship between entities requires data annotators to have deep domain knowledge accumulation and stable labor quality, which consumes a lot of man-hours. Summary of the invention

[0006] In order to solve the deficiencies in the prior art, the purpose of the present invention is to provide a distribution network dispatching knowledge fusion and graph storage method and system based on a graph database, so as to solve the technical bottlenecks of the existing distribution network dispatching knowledge system in the application of systematic knowledge linking, auxiliary decision-making reference query, etc.

[0007] The present invention adopts the following technical solution.

[0008] On the one hand, the present invention proposes a distribution network scheduling knowledge fusion and graph storage method based on a graph database, including:

[0009] Step 1: Collect semi-structured data and unstructured data of the distribution network dispatching system; store the semi-structured data and unstructured data as ontology in the graph database;

[0010] Step 2, extract triples of ontology in the graph database to obtain {first ontology and attributes, ontology relationship, second ontology and attributes} as the distribution network dispatching entity knowledge graph triple; take the first ontology and attributes, the second ontology and attributes in each distribution network dispatching entity knowledge graph triple as the first point and the second point respectively, and take the ontology relationship in each distribution network dispatching entity knowledge graph triple as the edge; use the first point, the second point and the edge to establish a subgraph model of the distribution network dispatching entity knowledge graph; use the subgraph model of the distribution network dispatching entity knowledge graph to express the distribution network dispatching entity, wherein the point of the distribution network dispatching entity knowledge graph subgraph model represents the distribution network dispatching entity, and the edge of the subgraph model of the distribution network dispatching entity knowledge graph represents the entity relationship between the distribution network dispatching entities;

[0011] The subgraph model of the knowledge graph of the distribution network scheduling entity is stored in the graph database;

[0012] Step 3, collecting event data of the distribution network dispatching system; storing the event data in a graph database;

[0013] Step 4, extract triples of event data in the graph database to obtain {first event, event relationship, second event} as the distribution network dispatch event knowledge graph triple; take the first event and the second event in each distribution network dispatch event knowledge graph triple as the first node and the second node respectively, and take the event relationship in each distribution network dispatch event knowledge graph triple as the directed edge; use the first node, the second node and the directed edge to establish a subgraph model of the distribution network dispatch event knowledge graph; the subgraph model of the distribution network dispatch event knowledge graph is a directed cyclic graph, and the distribution network dispatch event is expressed using the subgraph model of the distribution network dispatch event knowledge graph, and the nodes of the subgraph model of the distribution network dispatch event knowledge graph represent the distribution network dispatch event, and the directed edges of the subgraph model of the distribution network dispatch event knowledge graph represent the event relationship between the distribution network dispatch events;

[0014] The subgraph model of the knowledge graph of distribution network scheduling is stored in the graph database;

[0015] Step 5, using the knowledge fusion method to form a link path from the subgraph model of the distribution network dispatching knowledge graph to the subgraph model of the distribution network dispatching entity knowledge graph; based on multiple link paths, the knowledge fusion of the distribution network dispatching entity knowledge graph and the distribution network dispatching knowledge graph is realized;

[0016] Step 6: According to the field type, the distribution network scheduling entity knowledge graph and the distribution network scheduling principle knowledge graph with knowledge fusion are stored in the graph database.

[0017] Preferably, in step 1, the semi-structured data of the distribution network dispatching system includes: important power grid user information, power grid equipment topology connection information; the unstructured data of the distribution network dispatching system includes: dispatching plan text, historical alarm information records, risk records, risk warning notices, and historical operation records.

[0018] Preferably, in step 2, the subject includes: the subject of the equipment in the operation of the dispatching system, the subject of issuing and receiving instruction operations, the subject of place names, the subject of important users, and the subject of phenomenon names;

[0019] Among them, the information of the equipment subjects in the operation of the dispatching system includes: load, switch, power supply, connection line; the information of the subject that issues and receives the command operation includes: department name, operation team name, duty personnel name; the information of the place name subject includes: region name; the information of the important user subject includes: the name of the important power supply user; the information of the phenomenon name subject includes: fault type name, risk type name;

[0020] The attributes of the ontology include: number, level, time information, and status information;

[0021] Among them, the number information includes: plan number, fault number, user number, equipment number; the level information includes: fault level, risk level, equipment model; the time information includes: recording time, overload time; the status information includes: fixed-cut protection status, switch status, fault handling status;

[0022] Ontological relations include: hierarchical relations, action relations;

[0023] Among them, hierarchical relationships include: superior-subordinate relationships, parallel relationships; action relationships include: recording, patrolling, monitoring, and operation.

[0024] Preferably, in step 3, the event data of the distribution network dispatching system includes: dispatching plan text, historical alarm information records, risk records, risk warning notices, and historical operation records.

[0025] Preferably, in step 4, the event includes an event trigger word, an event argument element, and an attribute of the event;

[0026] Event trigger words include: fault action name, operation name; event argument elements include: event participant, event core action, event object, event occurrence time, event occurrence location; event attributes include: type, status.

[0027] Preferably, in step 4, the relationship between things includes: cause-effect relationship, conditional relationship, mutually exclusive relationship, composition relationship, transition relationship, and hierarchical relationship.

[0028] Preferably, step 5 comprises:

[0029] Step 5.1, according to the type of the entity to be associated in the distribution network scheduling knowledge graph, filter the entity knowledge list from the distribution network scheduling entity knowledge graph;

[0030] Step 5.2, based on the knowledge fusion method, the upstream information and downstream information of the entities to be associated in the distribution network dispatching knowledge graph are used as features to link the entities to be associated to the knowledge entity list; the entities of the distribution network dispatching knowledge graph are linked to the knowledge entities of the distribution network dispatching entity knowledge graph through multiple link channels.

[0031] Preferably, in step 5.2, the knowledge fusion method includes entity acquisition and entity linking, which are as follows:

[0032] Step 5.2.1, preprocessing and normalizing the data text describing the same entity in the distribution network dispatching theory knowledge graph and the distribution network dispatching entity knowledge graph, wherein the preprocessing includes the normalization of syntax and data;

[0033] Step 5.2.2, based on the block division method of proximity classification, multiple Map-Reduce operations are performed on the preprocessed and normalized data text to obtain multiple entity blocks, and the number of entities in each entity block is balanced;

[0034] Step 5.2.3, in any entity block, calculate the entity similarity between any two entities, where the entity similarity is represented by an attribute similarity function; and implement entity linking based on entity similarity;

[0035] Step 5.2.4, evaluate the entity linking results based on the evaluation indicators, end the entity linking when the evaluation indicators are met, and repeat steps 5.2.1 to 5.2.3 when the evaluation indicators are not met; wherein the evaluation indicators include: overall running time, accuracy, and recall rate.

[0036] Preferably, in step 6, the field types include floating point type, integer type, and string type.

[0037] On the other hand, the present invention proposes a distribution network scheduling knowledge fusion and graph storage system based on a graph database, and the system implements a distribution network scheduling knowledge fusion and graph storage method based on a graph database.

[0038] The beneficial effects of the present invention are as follows:

[0039] 1) According to the demand and current situation for the organization and storage of entity knowledge structure in the distribution network dispatching system, the present invention respectively establishes a subgraph model of the distribution network dispatching entity knowledge graph and a subgraph model of the distribution network dispatching factual knowledge graph; the modeling structure of each subgraph model is clear and efficient, which provides a basis for the subsequent modeling and storage of the distribution network dispatching operation knowledge system, and effectively improves the level of intelligence in the field of distribution network dispatching;

[0040] 2) Based on the characteristics of graph database, the subgraph model is defined, knowledge is integrated and stored; the knowledge integration of the distribution network dispatch entity knowledge graph and the distribution network dispatch event knowledge graph is realized, where the knowledge integration includes entity alignment, which can normalize entities with similar meanings in different graphs;

[0041] 3) When the present invention performs knowledge fusion, the entities to be associated are large data that do not have computational dependencies with each other. Multiple Map-Reduce operations are used for the entities to be associated, so that the number of entities in each entity block is balanced, which is convenient for parallel processing of data. When the number of entities in all entity blocks is equal, the performance quality and speed are improved through block operation; and through block operation, it is realized to select potential matching record pairs from all entity pairs in a given knowledge base as candidates, and the size of the candidates is reduced as much as possible;

[0042] 4) The present invention uses entity similarity to achieve entity alignment, thereby realizing entity linking of the distribution network scheduling knowledge graph and the distribution network scheduling entity knowledge graph, wherein entity similarity is used to determine whether two entities represent the same meaning. This method is convenient and easy to implement, and can improve the efficiency of knowledge fusion processing. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 It is a flowchart of the steps of the distribution network scheduling knowledge fusion and graph storage method based on graph database of the present invention. DETAILED DESCRIPTION

[0044] The present application is further described below in conjunction with the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and cannot be used to limit the protection scope of the present application.

[0045] like Figure 1 On the one hand, the present invention proposes a distribution network scheduling knowledge fusion and graph storage method based on a graph database, including steps 1 to 6.

[0046] Step 1: Collect semi-structured data and unstructured data of the distribution network dispatching system; store the semi-structured data and unstructured data as ontology in the graph database.

[0047] Specifically, in step 1, the semi-structured data of the distribution network dispatching system includes: important power grid user information, power grid equipment topology connection information; the unstructured data of the distribution network dispatching system includes but is not limited to: dispatching plan text, historical alarm information records, risk records, risk warning notices, and historical operation records.

[0048] Step 2, extract triples of ontology in the graph database to obtain {first ontology and attributes, ontology relationship, second ontology and attributes} as the distribution network dispatching entity knowledge graph triple; take the first ontology and attributes, the second ontology and attributes in each distribution network dispatching entity knowledge graph triple as the first point and the second point respectively, and take the ontology relationship in each distribution network dispatching entity knowledge graph triple as the edge; use the first point, the second point and the edge to establish a subgraph model of the distribution network dispatching entity knowledge graph; use the subgraph model of the distribution network dispatching entity knowledge graph to express the distribution network dispatching entity, wherein the point of the distribution network dispatching entity knowledge graph subgraph model represents the distribution network dispatching entity, and the edge of the subgraph model of the distribution network dispatching entity knowledge graph represents the entity relationship between the distribution network dispatching entities;

[0049] The subgraph model of the distribution network scheduling entity knowledge graph is stored in the graph database.

[0050] Specifically, in step 2, the subject includes: the subject of equipment in the operation of the dispatching system, the subject of issuing and receiving instruction operations, the subject of place names, the subject of important users, and the subject of phenomenon names;

[0051] Among them, the information of the equipment subjects in the operation of the dispatching system includes: load, switch, power supply, connection line; the information of the subject that issues and receives the command operation includes: department name, operation team name, duty personnel name; the information of the place name subject includes: region name; the information of the important user subject includes: the name of the important power supply user; the information of the phenomenon name subject includes: fault type name, risk type name;

[0052] The attributes of the ontology include: number, level, time information, and status information;

[0053] Among them, the number information includes: plan number, fault number, user number, equipment number; the level information includes: fault level, risk level, equipment model; the time information includes: recording time, overload time; the status information includes: fixed-cut protection status, switch status, fault handling status;

[0054] Ontological relations include: hierarchical relations, action relations;

[0055] Among them, hierarchical relationships include: superior-subordinate relationships, parallel relationships; action relationships include: recording, patrolling, monitoring, and operation.

[0056] In an embodiment of the present invention, the distribution network scheduling entity knowledge graph includes a triple structure of subject (attribute)-relationship-subject (attribute), forms a topological connection relationship of points and edges, and expresses the relationship between power grid entity information.

[0057] Step 3: Collect event data of the distribution network dispatching system; store the event data in the graph database.

[0058] Specifically, in step 3, the event data of the distribution network dispatching system includes: dispatching plan text, historical alarm information records, risk records, risk warning notices, and historical operation records.

[0059] Step 4, extract triples of event data in the graph database to obtain {first event, event relationship, second event} as the distribution network dispatch event knowledge graph triple; take the first event and the second event in each distribution network dispatch event knowledge graph triple as the first node and the second node respectively, and take the event relationship in each distribution network dispatch event knowledge graph triple as the directed edge; use the first node, the second node and the directed edge to establish a subgraph model of the distribution network dispatch event knowledge graph; the subgraph model of the distribution network dispatch event knowledge graph is a directed cyclic graph, and the distribution network dispatch event is expressed using the subgraph model of the distribution network dispatch event knowledge graph, and the nodes of the subgraph model of the distribution network dispatch event knowledge graph represent the distribution network dispatch event, and the directed edges of the subgraph model of the distribution network dispatch event knowledge graph represent the event relationship between the distribution network dispatch events;

[0060] The subgraph model of the distribution network scheduling knowledge graph is stored in the graph database.

[0061] Specifically, in step 4, the event includes event trigger words, event argument elements, and event attributes;

[0062] Event trigger words include: fault action name, operation name; event argument elements include: event participant, event core action, event object, event occurrence time, event occurrence location; event attributes include: type, status.

[0063] Specifically, in step 4, the relationship between things includes: cause and effect relationship, conditional relationship, mutually exclusive relationship, composition relationship, transition relationship, and hierarchical relationship.

[0064] In an embodiment of the present invention, the distribution network dispatching knowledge graph includes an event-relationship-event structure construction form, forming a directed cyclic graph, in which nodes represent events and directed edges represent logical relationships between events, namely, relationship between events, thereby expressing the relationship between power grid event information.

[0065] Step 5, using the knowledge fusion method to form a link path from the subgraph model of the distribution network dispatching knowledge graph to the subgraph model of the distribution network dispatching entity knowledge graph; based on multiple link paths, the knowledge fusion of the distribution network dispatching entity knowledge graph and the distribution network dispatching knowledge graph is realized.

[0066] Specifically, step 5 includes:

[0067] Step 5.1, according to the type of the entity to be associated in the distribution network scheduling knowledge graph, filter the entity knowledge list from the distribution network scheduling entity knowledge graph;

[0068] Step 5.2, based on the knowledge fusion method, the upstream information and downstream information of the entities to be associated in the distribution network dispatching knowledge graph are used as features to link the entities to be associated to the knowledge entity list; the entities of the distribution network dispatching knowledge graph are linked to the knowledge entities of the distribution network dispatching entity knowledge graph through multiple link channels.

[0069] Specifically, in step 5.2, the knowledge fusion method includes entity acquisition and entity linking, as follows:

[0070] Step 5.2.1, preprocessing and normalizing the data text describing the same entity in the distribution network dispatching theory knowledge graph and the distribution network dispatching entity knowledge graph, wherein the preprocessing includes the normalization of syntax and data;

[0071] In the embodiment of the present invention, since there are deviations in the way data texts describe the same entity, the data preprocessing method includes grammar and data normalization, and the normalized data text can be used for subsequent precise processing.

[0072] Step 5.2.2, based on the block division method of proximity classification, multiple Map-Reduce operations are performed on the preprocessed and normalized data text to obtain multiple entity blocks, and the number of entities in each entity block is balanced;

[0073] In the embodiment of the present invention, there are most entities that do not need to be connected in a given data text. By using the block division method of proximity classification, entities that potentially need to be matched can be selected, the number of candidate items can be reduced, and the coverage rate can be guaranteed. Load balancing after multiple Map-Reduce operations can balance the number of entities in each block after block division, ensuring the degree of performance improvement.

[0074] Step 5.2.3, in any entity block, calculate the entity similarity between any two entities, where the entity similarity is represented by an attribute similarity function; and implement entity linking based on entity similarity;

[0075] In the embodiment of the present invention, recording the similarity between two entities a and b includes recording the similarity of their attributes and the entity similarity obtained according to the attribute similarity.

[0076] Step 5.2.4, evaluate the entity linking results based on the evaluation indicators, end the entity linking when the evaluation indicators are met, and repeat steps 5.2.1 to 5.2.3 when the evaluation indicators are not met; wherein the evaluation indicators include: overall running time, accuracy, and recall rate.

[0077] Step 6: According to the field type, the distribution network scheduling entity knowledge graph and the distribution network scheduling principle knowledge graph with knowledge fusion are stored in the graph database.

[0078] Specifically, in step 6, the field types include floating point type, integer type, and string type. When performing a search index, integer type can usually be quickly located, so when storing digital type information such as numbers, integer type storage is preferred.

[0079] On the other hand, the present invention proposes a distribution network scheduling knowledge fusion and graph storage system based on a graph database, and the system implements a distribution network scheduling knowledge fusion and graph storage method based on a graph database.

[0080] The applicant of the present invention has made a detailed explanation and description of the implementation examples of the present invention in conjunction with the drawings in the specification. However, those skilled in the art should understand that the above implementation examples are only preferred implementation schemes of the present invention, and the detailed description is only to help readers better understand the spirit of the present invention, but not to limit the scope of protection of the present invention. On the contrary, any improvements or modifications based on the inventive spirit of the present invention should fall within the scope of protection of the present invention.

Claims

1. A distribution network scheduling knowledge fusion and graph storage method based on a graph database, characterized in that: The method comprises: Step 1: Collect semi-structured data and unstructured data of the distribution network dispatching system; store the semi-structured data and unstructured data as ontology in the graph database; Step 2, extract triples of ontology in the graph database to obtain {first ontology and attributes, ontology relationship, second ontology and attributes} as the distribution network dispatching entity knowledge graph triple; take the first ontology and attributes, the second ontology and attributes in each distribution network dispatching entity knowledge graph triple as the first point and the second point respectively, and take the ontology relationship in each distribution network dispatching entity knowledge graph triple as the edge; use the first point, the second point and the edge to establish a subgraph model of the distribution network dispatching entity knowledge graph; use the subgraph model of the distribution network dispatching entity knowledge graph to express the distribution network dispatching entity, wherein the point of the distribution network dispatching entity knowledge graph subgraph model represents the distribution network dispatching entity, and the edge of the subgraph model of the distribution network dispatching entity knowledge graph represents the entity relationship between the distribution network dispatching entities; The subgraph model of the knowledge graph of the distribution network scheduling entity is stored in the graph database; Step 3, collecting event data of the distribution network dispatching system; storing the event data in a graph database; Step 4, extract triples of event data in the graph database to obtain {first event, event relationship, second event} as the distribution network dispatch event knowledge graph triple; take the first event and the second event in each distribution network dispatch event knowledge graph triple as the first node and the second node respectively, and take the event relationship in each distribution network dispatch event knowledge graph triple as the directed edge; use the first node, the second node and the directed edge to establish a subgraph model of the distribution network dispatch event knowledge graph; the subgraph model of the distribution network dispatch event knowledge graph is a directed cyclic graph, and the distribution network dispatch event is expressed using the subgraph model of the distribution network dispatch event knowledge graph, and the nodes of the subgraph model of the distribution network dispatch event knowledge graph represent the distribution network dispatch event, and the directed edges of the subgraph model of the distribution network dispatch event knowledge graph represent the event relationship between the distribution network dispatch events; The subgraph model of the knowledge graph of distribution network scheduling is stored in the graph database; Step 5, using the knowledge fusion method to form a link path from the subgraph model of the distribution network dispatching knowledge graph to the subgraph model of the distribution network dispatching entity knowledge graph; based on multiple link paths, the knowledge fusion of the distribution network dispatching entity knowledge graph and the distribution network dispatching knowledge graph is realized; Step 5 includes: Step 5.1, according to the type of the entity to be associated in the distribution network dispatching knowledge graph, the entity knowledge list is screened from the distribution network dispatching entity knowledge graph; Step 5.2, based on the knowledge fusion method, the upstream information and downstream information of the entity to be associated in the distribution network dispatching knowledge graph are used as features to link the entity to be associated to the knowledge entity list; the entity of the distribution network dispatching knowledge graph is linked to the knowledge entity of the distribution network dispatching entity knowledge graph through multiple link channels; In step 5.2, the knowledge fusion method includes entity acquisition and entity linking, as follows: Step 5.2.1, preprocess and normalize the data text describing the same entity in the distribution network dispatching knowledge graph and the distribution network dispatching entity knowledge graph, wherein the preprocessing includes the normalization of syntax and data; Step 5.2.2, based on the block method of proximity classification, multiple Map-Reduce operations are performed on the preprocessed and normalized data text to obtain multiple entity blocks, and the number of entities in each entity block is balanced; Step 5.2.3, in any entity block, calculate the entity similarity between any two entities, wherein the entity similarity is represented by the attribute similarity function; realize entity linking based on entity similarity; Step 5.2.4, evaluate the entity linking results based on the evaluation index, terminate the entity linking when the evaluation index is met, and repeat steps 5.2.1 to 5.2.3 when the evaluation index is not met; wherein the evaluation index includes: overall running time, accuracy, and recall rate; Step 6: According to the field type, the distribution network scheduling entity knowledge graph and the distribution network scheduling principle knowledge graph with knowledge fusion are stored in the graph database.

2. The distribution network scheduling knowledge fusion and graph storage method based on graph database according to claim 1 is characterized in that: In step 1, the semi-structured data of the distribution network dispatching system includes: important power grid user information, power grid equipment topology connection information; the unstructured data of the distribution network dispatching system includes: dispatching plan text, historical alarm information records, risk records, risk warning notices, and historical operation records.

3. The distribution network scheduling knowledge fusion and graph storage method based on graph database according to claim 1 is characterized in that: In step 2, the subject includes: the subject of equipment in the operation of the dispatching system, the subject of issuing and receiving instruction operations, the subject of place names, the subject of important users, and the subject of phenomenon names; Among them, the information of the equipment subjects in the operation of the dispatching system includes: load, switch, power supply, connection line; the information of the subject that issues and receives the command operation includes: department name, operation team name, duty personnel name; the information of the place name subject includes: region name; the information of the important user subject includes: the name of the important power supply user; the information of the phenomenon name subject includes: fault type name, risk type name; The attributes of the ontology include: number, level, time information, and status information; Among them, the number information includes: plan number, fault number, user number, equipment number; the level information includes: fault level, risk level, equipment model; the time information includes: recording time, overload time; the status information includes: fixed-cut protection status, switch status, fault handling status; Ontological relations include: hierarchical relations, action relations; Among them, hierarchical relationships include: superior-subordinate relationships, parallel relationships; action relationships include: recording, patrolling, monitoring, and operation.

4. The distribution network scheduling knowledge fusion and graph storage method based on graph database according to claim 1 is characterized in that: In step 3, the event data of the distribution network dispatching system includes: dispatching plan text, historical alarm information records, risk records, risk warning notices, and historical operation records.

5. The method for distribution network scheduling knowledge fusion and graph storage based on graph database according to claim 1 is characterized in that: In step 4, the event includes event trigger words, event argument elements, and event attributes; Event trigger words include: fault action name, operation name; event argument elements include: event participant, event core action, event object, event occurrence time, event occurrence location; event attributes include: type, status.

6. The distribution network scheduling knowledge fusion and graph storage method based on graph database according to claim 1 is characterized in that: In step 4, the relationships between things include: cause and effect relationship, conditional relationship, mutually exclusive relationship, composition relationship, transitional relationship, and hierarchical relationship.

7. The distribution network scheduling knowledge fusion and graph storage method based on graph database according to claim 1 is characterized in that: In step 6, the field types include floating point, integer, and string.

8. A distribution network dispatching knowledge fusion and graph storage system based on a graph database, characterized in that: The system implements the distribution network scheduling knowledge fusion and graph storage method based on the graph database as described in any one of claims 1 to 7.

Images