Research on power grid fault diagnosis based on model matching and AI correction method with multi-dimensional data

A power grid fault diagnosis method based on model matching and AI correction method with multi-dimensional data is proposed to solve the problem that the mathematical model and operation of the current fault diagnosis method are complicated and the diagnosis result is difficult to explain. Firstly, the mathematical model of fault coding is established. Then, the method of correcting the misinformation of alarm information by discrete Hopjield neural network is introduced. Then the standard fault feature library of the single line is established by K-means clustering method, and the method of fault type identification is given. Finally, the actual grid fault is taken as an example to verify the feasibility of the proposed method.