Using the wavelet decomposition method for monitoring of amplitude-manipulated signals of railway automation

Abstract

In railway signaling systems, some parameters required for monitoring and controlling train movements are transmitted using amplitude shift keying (ASK) signals. ASK signals are more often used in railway automated systems due to their simpler generation by contact transmitters, but such signals have a low level of immunity and interference in them can lead to incorrect operation of railway signaling systems. The efficiency of using the signal decomposition method by wavelet transform to reveal distortions for continuous ASK railway signals have been investigated in the work. Simulated signals were generated as a pure ASK signals with a carrier frequency of 420 Hz and a modulation frequency of 8 Hz, as well as five variants of the original signal with disturbances in the form of signals with distorted signal shape, signals with swells, signals with sags, signals with 50 Hz interferences, and signals with sharp pulses . The signal segments of ten periods were decomposed using a five-level wavelet packet transform (WPT). For all investigated signals, the distortions or interferences in them lead, not only to changes in distribution of partial energies at the nodes, but also to the appearance of strong differences in the form of oscillations of the WPT coefficients. The variations in WPT coefficients in low-frequency nodes reflect changes in the waveform due to pulse distortion, sags, swells, etc. The obtained results confirm the efficiency of using the wavelet decomposition method in order to detect distortions in the ASK continuous railway signals.