ABSTRACT
The track defects detection method proposed in this paper emphasizes particularly on the analysis of the Bogie Acceleration (BA) measurements of the in-service vehicle. To validate the feasibility of the detection method, a vehicle-track dynamic model with dipped joint, thermite welds is built first by using SIMPACK. The BA collected by acceleration sensors installed on the four corners of bogie frame is analyzed to extract time-frequency features of the various track defects. The simulation results show that feature frequencies obtained from the Continuous Wavelets analysis (CW) and feature modes derived from the Empirical Mode Decomposition (EMD) can be used for track defects diagnosis. After that, field test is carried out at Shanghai metro Line 1. The results indicate that the proposed method can detect dipped joint and thermite weld defects with the aid of BA measurements from in-service train successfully.
