Application of Wavelet Transform and Fuzzy Pattern Recognition in Ultrasonic Detection

This paper studies application of wavelet transform and fuzzy pattern recognition to ultrasonic detection of bonding quality for thin composite plate. Firstly, ultrasonic NDT (Non Destructive Testing) detection method is introduced and the characteristic of the echo signal is discussed. Secondly, Wavelet transform is used to process echo signals; three characteristics of echo signal in time-frequency domain are extracted. Lastly, an algorithm of fuzzy pattern recognition based on the maximum membership degree principle is designed. Simulation proves that the algorithm can be used to recognize different echo signals quantitatively and effectively.