Signal processing, Wavelet, Underwater discharge, Time–frequency analysis, De-noising, Telecommunication, TK5101-6720, Electronics, TK7800-8360

Abstract The analysis of underwater discharge signals is of great significance for its application. Wavelet-based de-noising and analysis technology is an effective means to study underwater discharge signals. The selection of wavelets is the key to the accuracy of wavelet analysis. A scheme of wavelet selection is provided in this paper. Based on the signal characteristics and actual noise, the reference target signal and noisy signal are constructed to ensure the accuracy of wavelet performance evaluation. Cross-correlation coefficient, root mean square error, signal-to-noise ratio, and smoothness are chosen as evaluation indexes and fused by the coefficient of variation method. The selected optimal wavelet is used to process the underwater wire-guided discharge signals. The results show that the scheme is feasible and practical.