AUTHOR=Shuai Zhang , Qu Na , Zheng Tianfang , Hu Congqiang , Lu Senxiang TITLE=Research on arc fault detection using ResNet and gamma transform regularization JOURNAL=Frontiers in Energy Research VOLUME=Volume 11 - 2023 YEAR=2023 URL=https://www.frontiersin.org/journals/energy-research/articles/10.3389/fenrg.2023.1069119 DOI=10.3389/fenrg.2023.1069119 ISSN=2296-598X ABSTRACT=Series arc fault is the main cause of electrical fire in low-voltage distribution system. A fast and accurate detection system can reduce the risk of fire effectively. In this paper, series arc experiment is carried out for different kinds of electrical load. The time-domain current is analyzed by Morlet wavelet. Then, the multiscale wavelet coefficients are expressed as the coefficient matrix. In order to meet the data dimension requirements of neural networks, a color domain transformation method is used to transform the feature matrix into an image. A regularization method based on gamma transform is proposed for small sample data sets. The results show that this method can suppress the overfitting phenomenon of pre-training ResNet50 effectively, and the accuracy rate is increased to 96.53%. This method fuse fault features of 64 different scales, and provides a valuable manually labeled arc fault dataset. Compared with other typical lightweight networks, this method has the best detection performance.