Design And Implementation Of Knowledge Graph For Power Grid Equipment

Grid equipment is the core component of the power system.The status assessment,maintenance,testing and repair of grid equipment are the key links to ensure the reliable operation of the power system.These links are collectively referred to as the inspection of grid equipment.With the development of power technology,there are more and more types of grid equipment,and the equipment structure is becoming more and more complex,which makes the inspection of power grid equipment more and more difficult.Since the inspection work is highly dependent on the professional skills and work experience of the staff,the above problems have led to the continuous increase in the learning costs of the power grid staff,seriously affecting the efficiency of the inspection work.At the same time,with the development of data acquisition technology and the advancement of power grid informatization,the amount of text data about power grid equipment has surged.These texts contain a lot of valuable domain expertise,which can be used to build a knowledge base and provides auxiliary support functions to reduce the learning cost of the inspection skills.However,since most of the texts are unstructured texts,the professional knowledge in them cannot be directly utilized.In order to solve this problem,this paper designs and implements a knowledge graph system for power grid equipment based on the knowledge graph construction technology,which can store the professional knowledge in the form of knowledge graph,and provide knowledge analysis and reasoning capabilities based on the knowledge graph.First of all,this paper studies the construction method of domain knowledge graph,and constructs a large-scale knowledge graph dataset in the field of power grid equipment based on text data collection technology,named entity recognition technology and entity relationship extraction technology.This paper extracts knowledge based on 97,687 texts in the field of power grid equipment from multiple sources and with different structures.This paper extracts a total of 33,512 entities in 14 categories and 85,412 relationships in 12 categories,and constructs a knowledge graph in the field of power grid equipment.Compared with other knowledge graphs in the field,the knowledge graph in this paper has the characteristics of large scale,wide coverage and strong domain characteristics.Secondly,this paper proposes a dynamic update method for knowledge graph.This method dynamically updates the data in the knowledge graph through six upgrade executors to continuously supplement new domain knowledge.Based on this method,the version management of knowledge graph data changes can be realized,and the details of each data update event will be recorded in the log system,then a data rollback mechanism is provided.In addition,the knowledge graph dynamic update method is also highly scalable.When the number of knowledge graph data upgrade method increases,only the new corresponding upgrade executors need to be added.Thirdly,this paper proposes an open-domain zero-shot knowledge graph completion model.This model is based on the reasoning ability of the knowledge graph itself and the design of generative confrontation ideas.In the absence of a large-scale training data set,this model can learn the relation features from the natural language text description of the new relation,and complete the relation complement in the low-dimensional vector space of the knowledge graph.It solves the problem of missing links in the knowledge graph in the field of power grid equipment.Finally,this paper designs and implements a knowledge graph system for power grid equipment.Based on text data and knowledge graph data in the field of power grid equipment,the system provides domain data management,full-text search,similar text recommendation,and data visualization functions.Through a series of system tests,it is verified that the function and performance of the system meet the design expectations,and it can play a powerful support role in the inspection work of power grid equipment.