The design and testing of an Electronic Nose prototype for classification problem

Abstract

This paper reports the design of an Electronic Nose prototype based on commercially available metal oxide gas sensors and a temperature sensor. The Electronic Nose system comprises of a sampling chamber, sensing chamber, pumps, data acquisition system and controller unit and a computer for data analysis. To verify its repeatability, reproducibility and discriminative characteristics, several experiments were conducted using three beverages: blackcurrant juice, orange juice and soy milk. Results obtained from the experiments verify that the developed Electronic Nose is reliable to produce correct measurement with high repeatability, reproducibility and discriminative ability. © 2010 IEEE.