Investigating the feasibility of using principal component analysis for ultrasonic classification of gas mixtures

Abstract

On-line measurement of the energy content of natural gas is of interest for both industry and customers, since the energy content determines its monetary value. Experiments with pulsed ultrasound show that, in addition to changes in speed of sound and acoustic attenuation, there is also a change in the shape of the sound waveform. In this paper, we investigate the feasibility of using principal component analysis (PCA) to quantify this change in pulse shape. The principle is evaluated for pure oxygen, pure ethane, and mixtures of the two, for different pressures. The results show that by using PCA, it is possible to distinguish between pulses that have propagated through oxygen from pulses in ethane and mixtures of the two.