The research on power grid fault diagnosis is reviewed comprehensively. The sources and characteristics of its data

and the ways and importance of data preprocessing are analyzed. The principle

features and shortcomings of various intelligent diagnosis algorithms are briefly treated

the improvement of research results in recent years are discussed and future study directions are proposed. Then according to engineering practice

the architectural design in the online diagnosis system is summarized. Finally

the problem of power grid fault diagnosis in the background of big data in power grid dispatch and control is discussed

while the development trends are elucidated in the new situation.