METHOD AND SYSTEM FOR CLASSIFYING TYPE AND SEVERITY OF DEFECT IN WELDED PART

摘要

PROBLEM TO BE SOLVED: To provide a method and system for detecting a defect in a weld to determine the type and seriousness of the detected defect. SOLUTION: In the method for determining the type of a defect in a weld, the ultrasonic response signals gathered from a plurality of the measuring places set along the weld are analyzed to determine a defect position. The defect signal corresponding thereto and a plurality of the defect approach signals corresponding to the defect signal and the ultrasonic response signals from the measuring places on both sides of the defect position, are inputted to an educated artificial neural network. The educated artificial neural network discriminates the type of the defect positioned in the defect position on the basis of the defect signal and a plurality of the defect approach signals to output the type of the defect positioned in the defect position. The educated artificial neural network determines the classification of the degree of the defect on the basis of the defect signal and a plurality of the defect approach signals, and outputs them. COPYRIGHT: (C)2011,JPO&INPIT