摘要: |
For centuries, spikes, with marginal changes in terms of geometry and material, have been consistently providing reliable restrictions to rails. With the increasing axle load and operational speed, spikes are subjected to more demanding loading conditions, especially in the territories where tracks have high curvature. Because the cracks are typically underneath the spike head, it is very difficult to distinguish the broken spikes without physical examination, which causes formidable challenges in track health evaluation and operational safety. The objective of this research effort is trying to develop a low-cost, non-destructive and contact-free intelligent inspection system to identify broken spikes at the real-time rate. The proposed system will integrate laser excitation, acoustic analysis, computer vision and pattern recognition, and artificial intelligence (AI). The success of this research may significantly improve the efficiency and accuracy in spike inspection and enhance railroad track safety. |