Research on Optimization of Power Safety Intelligent Q&A System Based on Multi-source Knowledge Graph Enhanced Large Language Modeling

Abstract

 This paper focuses on the optimization research of intelligent question and answer system for electric power safety. Power safety is of great significance to the economy and life, but the existing intelligent Q&A system has shortcomings such as inaccurate answers and limited ability to deal with complex problems. The study analyzes the existing system architecture, functions, application scenarios and problems, and analyzes the multi-source knowledge graph and large language model technology, and finds that the combination of the two has a synergistic effect. To this end, optimization strategies are designed, including knowledge enhancement based on the knowledge graph, the supplementation of the knowledge graph by the large language model, and the combination of loosely-coupled and tightly-coupled system integration and intelligent interaction mechanisms. The optimized system can enhance fault diagnosis and preventive maintenance capability in power operation and maintenance, promote talent training and enrich the teaching form in power safety education, and also has significant economic and social benefits, which can help the intelligent development of the electric power industry and provide new ideas for the construction of intelligent question and answer system for power safety.