Vehicle - Track Coupling Modeling And Track Surface Depression Detection

Squats are a type of the most devastating track short-wave irregularity, which refers to the similar lobe-shaped local depression and arc crack defect. Real-time detection of the squats in railway infrastructure has a great significance for the safety of railway operation and impronement of the railway operation level. On the basis of absorbing and summarizing predecessors’research results, this paper demonstrates how squats are detected using axle box acceleration(ABA) measurements on trains.Firstly, the basic principle of detection of squats are introduced, as well as the mechanism and physical fitting model of squats. Theoretical derivation is made to obtain the equal relation between track irregularity frequency and the vertical acceleration of the axle box.Secondly, a coupling dynamics model of vehicle-track interactions is developed. The model consists of vehicle subsystem, track subsystem and wheel-rail coupling submodel. With harmonic track irregularity as model excitation, ABA is obtained. Comparing the analysis results of FFT with the theoretical value, the rationality and accuracy of the model are validated.Thirdly, an online detection of squats using Wavelets is proposed. This method collects vertical and horizontal ABA. Wavelet power spectrum, time wavelet energy spectrum and scale wavelet energy spectrum are respectively applied to the signals. The scale-averaged wavelet power is obtained according to the range of fault characteristic frequency, and squats are detected automaticly by the threshold. Experiments based on the vertical and longitudinal ABA are carried out.Besides, the signal-processing solution for the reduction of disturbances from wheel defects.The actual signal of a corporation proved the effencicy the proposed method.Finally, the length detection of squats based on Adaptive Morphological Filtering(AMF) and Frequency Slice Wavelet Transform(FSWT) is demonstrated.The basic principle of AMF and FSWT is introduced and the Frequency Slice Function(FSF) is modified. The analysis results of simulated and measured signals shown that the proposed method is capable of assessing the length of squats.This offered a new technical approach for railway track defects diagnosis.