Spectral-Time Methods for Analysis and Detection of Vehicle Noise in Geomonitoring Tasks

The problem of passive geophysical monitoring of vehicle objects based on their seismic-acoustic noise is considered. Heavy types of vehicles are considered as sources of vehicle oscillations: electric trains, freight trains, wheeled and tracked vehicles. The statement of the monitoring problem is given. Solving it includes a number of stages. The initial stage of monitoring is associated with the detection of moving vehicles against the background of external noise. The paper proposes two ways to increase the detection range: taking into account the characteristic narrowband frequency components in the oscillation spectrum in one method and broadband ones in the other. The solution is based on combining methods of spectral-time analysis and narrowband filtering of recorded oscillations. The results of the analysis of records of vehicle vibrations obtained in the field are presented. The features of the spectra of seismic and acoustic noise of various types of vehicle are considered. It has been shown that the predominant frequencies of seismic and acoustic oscillations arising during vehicle movement are in the low frequency range. The dependence of the characteristic narrowband frequency component on the speed mode of the vehicle is analyzed.