Magnetic Adsorption Non-destructive Testing Vehicle Based on TOFD and Rainflow Counting by Neural Network Technology

We innovatively applied neural network technology to the field of non-destructive testing of steel pipelines. According to the survey, there has been a serious gap in the development of intelligence and automation in the field of non-destructive testing of steel pipelines for a long time. The defect detection of the surface of the pipelines for natural gas and oil and wind turbine main poles is conducted manually with problems of low detection accuracy, high risk, and high work intensity. Manual work is easy to miss pipeline damage, causing safety accidents and leaving a safety hazard for production. Pipeline defect detection requires several hours of high-intensity work, and defect detection workers have huge work intensity. The proposed system automatically performs non-destructive defect detection on pipelines. The cloud platform displays ultrasonic images of defect detection in real time, automatically analyzes and saves records of the recognition results during the process, and finally, a defect detection report is generated on the cloud platform.