Study on the Wavelet Denoise and Characteristic Extraction of Echo Signals for Flaw Classification in Ultrasonic Testing

摘要

There are usually two problems in flaw classification in ultrasonic testing. One is noise in echo signal in ultrasonic testing, which is sometimes very difficult to eliminate; the other is criterion problem for validity evaluation in echo signal characteristics extraction. This paper presents a noise eliminating method for ultrasonic signal with wavelet denoise and makes use of sort separability criterion for the second problem. And these two methods are verified by some experiments. Firstly wavelet transform was used in denoising process of ultrasonic signal, then sort separability criterion was used to evaluate the characteristics choice of flaw signals, and finally the characteristics values of flaws in demodulated signal were classified by RBF neural network to validate the above methods. Result of the experiments show that due to wavelet denoise algorithm making best use of the information of time and frequency domain at the same time in ultrasonic echo signals, not only the denoise effect is obvious, but also flaws location is accurate; sort separability criterion takes the quantification measure effect on the characteristics extracting of flaw signals.