Developments and applications of electronic nose systems for gas mixtures classification and concentration estimation

Abstract

In this paper, electronic nose systems consisting of five low cost gas sensors and three auxiliary sensors are described. The devices are effectively applied to gases mixtures classification in refinery environment and in monitoring of patients’ breath on haemodialysis treatment. The systems exploit a classification algorithm based on support vector machine method and a least square regression model for concentration estimation. In particular, in the present work, the systems implementation and the results obtained during data acquisition and post-processing phases are reported.