1.School of Electrical Engineering, Beijing Jiaotong University, Beijing 100044, China;2.China Electric Power Research Institute, Beijing 100192, China

Explainable artificial intelligence (XAI), as a new type of artificial intelligence (AI) technologies, can present the logic of the AI process, reveal the AI black-box knowledge, and improve the credibility of the AI results. The deep coupling between XAI and power systems may accelerate the AI technology application in the power system and assist with the safety and stability of human-machine interaction. Therefore, this paper reviews the historical context, development needs, and hot technologies of XAI in the power system, summarizes its applications in source-load forecasting, operation control, fault diagnosis, and electricity market, and explores the application prospects of XAI in the power system around aspects such as interpretability, iterative framework, and number-matrix fusion. This paper aims to provide theoretical references and practical ideas for promoting the intelligent transformation and iterative human-machine interaction of the power system.