This paper reviews the intelligence methods that are applied widely for fault diagnosis of power networks, including expert system, artificial neural network, Bayesian network, optimization methods, support vector machine, fuzzy set theory, Petri net, information fusion of multi-data resources and multi-agent system. Their basic concepts and research status in power fault diagnosis domain are introduced. Their characteristics, disadvantages and development trends are elaborated from the aspects of practicability. Finally, this paper explores the development trend in future in the field of fault diagnosis according to the current problems in the field of grid fault diagnosis.