Based On The Generalized Cross-correlation Lms In The Application Of The Ultrasonic Signal Processing

The dissertation improves the precision by using the generalized cross-correlationalgorithm based on the LMS (Least Mean Square) algorithm instead of the traditionalcross-correlation algorithm， and verifys its good results through simulation andexperiment. It also uses traditional the smooth filter and the wavelet denoising methodto process the ultrasonic signal， and carries on the contrast with the LMS algorithm.The dissertation makes a detailed introduction of the non-destructive testing. Italso elaborates the ultrasonic testing situation at home and abroad and its status in thenon-destructive testing area. The signal gathered in the experiment includes massivenoises. These will have the serious influence to the experimental result. So how todenoise the signal is the first question. The LMS denoising method makes the contrastwith the traditional smooth filter and wavelet denoising method， in order to knowabout the effect of the improved algorithm. On one hand the LMS algorithm is used todenoise the ultrasonic signal gathered by the oscilloscope， on the other hand itcalculates the time delay by the cross-correlation algorithm.Firstly， the simulation is processed in the MATLAB. The random noise is carriedinto the ideal ultrasonic signal artificially. It takes the LMS algorithm to denoise thesimulated ultrasonic signal， and calculates the signal-to-noise ratio before and after thedenoising. At the same time it uses the traditional smooth filter and the waveletdenoising method to denoise the same simulated ultrasonic signal， and calculates thesignal-to-noise ratio，which is used as the contrast experiment with the LMS algorithm.Next， the experiment takes the CTS8077PR(ultrasonic pulse sending andreceiving instrument) as the hardware platform， and processes the ultrasonic signal inthe MATLAB. The real ultrasonic signal gathered by the oscilloscope is denoised by theLMS algorithm. At the same time the dissertation takes the traditional smooth filter andthe wavelet denoising method to denoise the real ultrasonic signal， and calculates thesignal-to-noise ratio.