Efficiency Enhancement of Electronic Nose Classification Using Artificial Neural Network

Abstract

Electronic nose is recently popular and has been used in various applications according to it is very facility and easy to use. This paper presents the classification method for personal shirt odor. The metal oxide gas sensors have been measured the shirt odor because it is stable, reliable and convenience to design the circuit to measure the VOCs from sample shirt. The evaluate highest signal changing value of gas sensor is then analyze using artificial neural network to classify the shirt odor of each person. The results show that electronic nose in this experiment can exactly classify the shirt odor of 7 volunteers. The relative absolute correctly is 99% that result is better than MDA and PCA technique.