Analysis of Electricity Safety in Scientific Research and Production Sites: A Novel HMM-VA-based Knowledge Graph Approach

The electricity load in scientific research and production sites is large, and the impact of electrical safety accidents is difficult to estimate. It is crucial to analyze the potential safety hazards and causes of safety accidents in these sites. Most of the traditional methods are generally realized through gradual troubleshooting of low-voltage terminal distribution system, power quality of electrical equipment, grounding system, but these methods are difficult to quickly and accurately determine the characteristics of hidden trouble. For this reason, this paper proposes an analysis method of potential safety hazards of power consumption in scientific research and production sites based on knowledge graph. Firstly, this paper extracts the unstructured security risk data and constructs an elastic search (ES) to store the risk data. Secondly, we utilized the Chineseword segmentation model based on self-attention mechanism-guided by syntactic dependency (AM-GSD) for word segmentation training on data within the engine and implemented the annotation of hidden entity word segmentation. Finally, this paper uses Neo4j Graph database to build a knowledge graph of power safety hazards, and realizes the analysis of power safety hazards in scientific research and production sites. The method proposed in this paper has been tested and verified by the actual data of safe power utilization in a scientific research site. The experimental results show that the proposed method can effectively display and retrieve hidden danger data.