Real time fault detection of railway vehicles and tracks

Abstract

This paper describes methods for fault detection of railway vehicles and tracks. We examine fault detection of tracks using multi-resolution analysis into on-board measurement data as model-free approach. For fault detection of vehicles, we adapt interacting multiple-model (IMM) algorithm as model-based approach. We carried out simulation studies and field tests with actual vehicle. The results suggest that track and vehicle fault can be effectively detected by proposed method.