Research of laser ultrasonic defect statistics recognition technology based on radial basis function neural network

Abstract

In this study, we detected the depth of surface defect at a statistic on rate of 95.8% or more by using the Radial Basis Function Neural Network, a Q-switched and pulsed Nd:YAG laser and optical fiber Fizeau interferometer. In the experiment, Nd:YAG pulsed laser was used for generating ultrasound in a sample, while an optical fiber Fizeau interferometer was used to detect the ultrasound generated by a laser line source. Our results showed that this system has a high sensitivity as well as a high spatial resolution, what is more important is that this system is applicable to practice field.