Reduction of Doppler effect for the needs of wayside condition monitoring system of railway vehicles

Technology of acoustic condition monitoring of vehicles in motion is based on the assumption that diagnostically relevant information is stored in the acoustic signal generated by a passing vehicle. Analyzing the possibilities of increasing the effectiveness of condition monitoring of a passing vehicle with stationary microphones, it should be noted that the acoustic signal recorded in these conditions is disturbed with the disturbance resulting from the Doppler effect. Reduction of signal's frequential structure disturbance resulting from the Doppler effect allows efficient analysis of changes in frequential structure of recorded signals and as a result extraction of relevant diagnostic information related with technical condition of running gear of vehicle. This article presents a method for removal of signal's frequential structure disturbances related with relative move of vehicles and stationary monitoring station. For elimination of the frequential non-stationary of signals disturbance-oriented dynamic signal resampling method was used. The paper provides a test of two methods for defining the time course of local disturbance of signal's frequential structure: the method based on the Hilbert transform and the method of analytical description of signal's disturbance based on the knowledge of a phenomenon that causes frequential non-stationarity of signals. As an example, the results of the processing and analysis of acoustic signals recorded by wayside measuring station, during the passage of WM-15A railway vehicle on an experimental track of Polish Railway Institute, are presented. (C) 2012 Elsevier Ltd. All rights reserved.