Intelligent recognition method for defects in friction stir welding based on ultrasonic phased array and related equipment

Abstract

The application discloses a kind of based on ultrasonic phased array's stir friction welding defect intelligent identification method and related equipment, method includes: obtaining the stir friction welding detection data of ultrasonic phased array, according to the stir friction welding detection data first image is constructed;The first image is input to the trained prediction model, and the defect information of stir friction welding is obtained;The defect includes any one or more in porosity, incomplete penetration, tunnel type hole or crack;The defect information includes any one or more in defect echo amplitude, defect height or defect length;The prediction model is based on YOLO big data model training obtains.This embodiment can automatically detect stir friction welding and improve the accuracy of identifying stir friction welding defect.The present application can be widely applied in nondestructive testing technical field.

Images