Abstract: In order to ensure the safe operation of distribution lines and improve power supply reliability, a short-circuit grounding fault detection method for distribution lines based on improved wavelet analysis is proposed. Combining improved wavelet analysis technology, high-precision sensors are used to collect electrical signals of distribution lines, and the signals are decomposed and reconstructed for processing. Based on the reconstructed signal mentioned above, extract short-circuit grounding fault features from the reconstructed signal. Using the extracted short circuit and ground fault feature vectors to determine the absolute and relative errors of distribution line operation, and then outputting the identification results of short circuit ground faults. The experimental results show that the short-circuit grounding fault detection method for distribution lines based on improved wavelet analysis has high detection accuracy and good real-time performance. Compared with the other two methods, this method has significant improvements in various performance indicators.