Continuous leakage location in noisy environment using modal and wavelet analysis with one AE sensor

Abstract

A noble method for continuous leakage source location with one sensor in gas-filled pipe and noisy environment is proposed based on wavelet analysis and modal location theory. The leakage signals were analyzed into high and low frequencies by wavelet decomposition and noises and reflected waves were omitted. Then the processed signals were reconstructed. A frequency range of 0–250kHz was selected using wavelet packet decomposition to analyze the flexural and extensional modes of leakage signals. In this frequency range, the peak magnitudes of wavelet transform could determine the arrival times of different modes. The frequencies which the peak magnitudes of wavelet transform occurred per them were chosen to calculate the group velocities. Experiments were carried out and the locating results show the high precision of this algorithm.