Rapid positioning test method for cylindrical structure penetration based on machine learning

Abstract Focusing on the requirements of rapid damage location test for multi-layer steel plates penetration of cylindrical structure in the complex and harsh environment, this paper studies the propagation characteristics of high-speed impact stress waves in steel plates. Also, the measurement effects of impact stress in the near zone of penetration hole are compared with different sensors, so as to select appropriate PZT sensors to measure the velocity of impact stress wave in the actual steel plates. Then the array layout scheme is designed, and the corresponding damage location algorithm is determined. Finally, the deep learning algorithm is used to improve the accuracy and efficiency of damage location.