Spectral Processing of Acoustic Signals for Pipeline Security Systems

Abstract

this paper is an overview of the existing pipeline security systems preventing unauthorized access to the pipe. It is shown that the most suitable solution is an acoustic sensor and optoelectronic recognition system. It operates as follows: the acoustic sensor is activated – the acoustic algorithm confirms the alarm – the optoelectronic system confirms the alarm – the alarm is triggered. The direct Fourier transform is used for acoustic signal recognition to identify the frequency spectrum of acoustic signals and compare it with the reference one. The above-mentioned direct transform includes a decomposition of the total acoustic signal from the sensor into original sinusoids. The purpose of this study is to upgrade the oil pipeline security systems for better efficiency. The proposed upgrade method unambiguously identifies any unauthorized access by analyzing and comparing the audio signal with the reference signal from a database of pre-recorded sounds to protect oil and gas pipelines. The solution also minimizes the false alarm rate.