Research on Fault Arc Detection Method Based on FOA-GRNN

The series fault arc caused by aging and damage of line insulation and poor electrical contact seriously threatens the power safety of low-voltage distribution system. Its characteristics such as low current, high temperature and strong concealment make it difficult to detect and identify. Based on this, a fault arc detection method based on wavelet features and generalized regression neural network is proposed. By building the experimental platform of fault arc, the current signals under typical resistive and inductive loads are collected, and the wave transform is used to extract the characteristics of the current signals. The training set and test set are constructed. The generalized regression neural network is optimized by the Drosophila optimization algorithm, and the smoothing factor of GRNN is optimized, so as to improve the recognition ability of the model and the smoothing factor of GRNN is optimized. Experimental results show that FOA-GRNN has better recognition accuracy and recognition rate than traditional BP network.