Bonding steel reinforcing adhesive layer defect detection method and device based on machine learning

• Main Authors: LIU ZEJIA, ZHOU NANJIE, LIU YONGFENG, GUO ZEMIAN, CHEN HONGZHE, ZHOU LICHENG, XIA GUIRAN
• Format: Patent
• Language: Chinese; English
• Published: 23.09.2022
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; ELECTRIC DIGITAL DATA PROCESSING; HANDLING RECORD CARRIERS; PHYSICS; PRESENTATION OF DATA; RECOGNITION OF DATA; RECORD CARRIERS

The invention discloses a machine learning-based sticky steel reinforcing adhesive layer defect detection method and device, and the method comprises the steps: constructing a sticky steel reinforcing adhesive layer defect model library, obtaining an ultrasonic reflection time-history response according to the sticky steel reinforcing adhesive layer defect model library, obtaining a training set and a test set according to the ultrasonic reflection time-history response, and carrying out the detection of the sticky steel reinforcing adhesive layer defect model library. The training set and the test set are sent to a deep learning model for deep learning, a sticky steel reinforcing adhesive layer defect detection model is obtained, ultrasonic reflection time history response corresponding to the sticky steel reinforcing adhesive layer defect to be detected is input into the sticky steel reinforcing adhesive layer defect detection model, and therefore sticky steel reinforcing adhesive layer defect detection is