An adaptive wavelet denoising method for the measuring system of EMP signals

Abstract

An adaptive wavelet denoising method is proposed to eliminate the noise induced by the measuring system of EMP signals. This method employs a threshold-searching strategy to select an optimal denoising threshold for a given system. Wavelet decomposition and reconstruction are combined with neural network nonlinear threshold-filtering units in the new denoising algorithm. Based on a group of training signal, the denoising threshold can be learned adaptively. The training algorithm and application examples are given in this paper.