Slection of wavelet fuction on denoise qulities of ultrasonic inspection signal

Abstract

It is suggested in this paper to denoise the ultrasonic inspection signals using wavelet transform. The theoretic base is that wavelet transform is a band-pass filter. Firstly, the frequencies of the inspection signal are evaluated according to the excited signal. Secondly, an appropriate scale for wavelet transform is computed following the laws between the scale and the frequency. Thirdly, wavelet transform with the scale would be operated. As results, the echoes by the flaws could be discerned in the wavelet coefficients. Finally, the propagation of the guided wave in a damaged pipe is simulated by code ANSYS and the signal processing are operated by code MATLAB. The different wavelet functions are also discussed aiming at the denoised qualities. The signal-to-noise ratio (SNR) is used to evaluate the quality for different wavelet. It could be concluded that the Morlet wavelet is suitable to operate the ultrasonic inspection signal excited by a sinusoidal signal modulation by Hanning windows.