Artificial Intelligence Approaches to Fault Diagnosis in Power Grids: A Review

In light of heightened interest in Artificial Intelligence (AI) applications in power systems, following the massive success of AlphaGo, this paper presents a comprehensive review of AI approaches to fault diagnosis (FD) in power grids, including expert system, artificial neural network, Petri network, fuzzy set theory, rough set theory, multi-agent system, and information fusion technology. A framework of AI applications in power system FD is first presented. Then the basic concepts, characteristics and practical applications of AI methods in the power grid FD are analyzed including their advantages and deficiencies. Finally, future trends and outlooks using AI in power grid fault detection and identification are discussed to promote further research in this field.