基于激光超声波金属机械零件表面缺陷检测方法

Surface Defect Detection Method of Metal Mechanical Parts Based on Laser Ultrasonic

针对目前金属表面缺陷检测时存在数据集不足、检测精度低的缺陷,提出了一种基于激光超声波金属机械零件表面缺陷检测方法。首先,研究了基于热弹性机制的激光超声生成机制,提出应用数值模拟产生大量超声波样本数据。其次,设计了一种基于深度学习的金属表面缺陷检测模型,从而实现端对端的金属表面缺陷检测。实验结果表明,与CNN和LSTM网络相比,所提金属表面检测模型综合性能更优。仿真结果进一步验证了所提模型对激光超声波下金属零件表面缺陷检测研究具有一定借鉴作用。

Aiming at the defects of insufficient data set and low detection accuracy in the current metal surface defect detection,a surface defect detection method of metal mechanical parts based on laser ultrasonic is proposed.Firstly,the mechanism of laser ultrasonic generation based on thermoelastic mechanism is studied,and a large number of ultrasonic sample data are generated by numerical simulation.Secondly,a metal surface defect detection model based on deep learning is designed to realize end-to-end metal surface defect detection.The experimental results show that the proposed metal surface detection model has better comprehensive performance than CNN and LSTM networks.The simulation results further verify that the proposed model has a certain reference value for the research of surface defect detection of metal parts under laser ultrasound.