Ultrasonic Flaw Detection Based On Wavelet Transform

Ultrasonic testing technology is widely used in industry, construction, transportation and other fields.This paper first presents the current research situation and develop tendency of ultrasonic testing and wavelet analysis at home and abroad. On the basis of this, the paper expound the fundamental theory of wavelet transform and Fourier transform, the classification and the physical parameters of the ultrasonic wave and several kinds of commonly used wavelet basis function. Then on the basis of pulse ultrasonic detecting theory, we bulid the hardware framework of the ultrasonic testing system. In accordance with various modules functional requirements, the relevant hardware circuits such as ultrasonic transmit, programmable gain amplifier and high speed data acquisition are designed. The test results show that this system design is correct and feasible.However, the flaw signals detected by this ultrasonic detection system are often mixed up with a large quantity of noise, affecting the correct judgment. So the study of some effective de-noising methods have very important significance. This article uses the method of wavelet threshold to deal with the noise signal. Through a lot experiments, we determine the parameters such as threshold rules, wavelet basis function and decomposition layers. When the wavelet coefficients of noise and signal have a large difference, we have a wide threshold selection space. On the basis of this theory, this paper put forward a kind of wavelet coefficient threshold pretreatment method, and after pretreatment we can larger the difference between the wavelet coefficients of noise and signal, which provide a prepare word for de-noising.Determine the threshold function of wavelet threshold de-noising method is an important step. The traditional threshold functions including soft threshold, hard threshold function and semi-soft threshold function. This paper analyze the advantages and disadvantages the mentioned threshold functions. From the signal to noise ratio and the mean square error (mse) point of view, compare with the traditional thresholding function, the proposed improved threshold method has a better de-noising effect. The experimental results show that the effectiveness of the improved threshold function is correct and feasible.