NEURAL-NETWORK-BASED ELECTRONIC NOSE FOR THE CLASSIFICATION OF AROMATIC SPECIES

In this work, an aroma identification system has been developed. Based on an array of semiconductor tin dioxide gas sensors and neural netwo rk processing algorithms, the system has proven a 100% success rate in the discrimination of five different aromatic species. An initial nin e sensor array was simplified to seven after a PCA analysis detected r edundancy between three of the sensors. Data processing and classifica tion performed by a feedforward artificial neural network with a hidde n layer and trained with a backpropagation algorithm showed no signifi cant performance differences between the complete and reduced sensor a rray which confirms the redundancy detected by the PCA analysis. Our r esults show that a reliable Electronic Nose system can be designed usi ng inexpensive and poorly selective chemical semiconductor gas sensors .