Research on Fault Diagnosis Method of Distributed Power Distribution Network Based on HHT and CNN

Accurate and reliable fault diagnosis is very important for improving relay protection performance and ensuring safe and stable operation of power system. This paper presents a fault diagnosis method for distributed power distribution network based on convolutional neural network. When fault occurs in the distribution network, FTU acquisition unit is used to sample the fault current and voltage signals of the distribution network. Hilbert-Huang transform is applied to the fault signal, and a two-dimensional time-frequency energy matrix is constructed. Then the time-frequency energy matrix is input into the convolutional neural network to realize the fault type identification of transmission lines in the distribution network with distributed power supply. The simulation results show that the model in this paper improves the speed of fault diagnosis, makes the model more robust, and can determine and judge fault types more accurately and quickly.