Welding defect signal detection based on particle swarm optimized ICA algorithm

摘要

In the Ultrasonic detection of welding defects, the defect echo signal carries the important feature information of sampling object. To solve the problem of accurate extraction of defect signal in the Ultrasonic detection of welding defects, the ICA algorithm based on Particle Swarm Optimization is proposed to extract the ultrasonic signal of welding defects, and using this algorithm to denoising processing of weld defects in practice. Experiment result show that the algorithm gets better stability and performance of separation of signal and noise than FastICA algorithm.