Application of the Ultrasonic Guided Wave Technique Based on PSO-ELM Algorithm in the Rail Fatigue Crack Assessment

Abstract Rail safety is very important, and fatigue cracking is one of the important factors affecting rail safety. Therefore, it is an urgent need to develop a safe and effective rail fatigue crack detection technology. Ultrasonic guided wave technology plays an important role in rail detection because of its long propagation distance and small attenuation. In order to realize the quantitative detection of rail fatigue crack, an ultrasonic guided wave technology based on particle swarm optimization–extreme learning machine (PSO-ELM) algorithm for evaluating the rail fatigue crack depth is proposed. The finite element method is used to establish the ultrasonic guided wave model in the rail, and the rail fatigue crack at different depths is simulated. The ultrasonic guided wave selected through the time window function of the excitation signal is used for analysis, and then nine features such as the time domain and the frequency domain of the ultrasonic guided wave are extracted. The PSO-ELM algorithm is used to identify the rail fatigue crack with different depths, and an ultrasonic guided wave-based detection system for the rail fatigue crack is built to verify the relevant theoretical results. The results of finite element simulation and the experiment show that ultrasonic guided wave technology based on PSO-ELM algorithm proposed can quantitatively evaluate the rail fatigue crack with different depths, with an accuracy of more than 99.95 %, which provides an effective method for the rail fatigue crack detection.