Gas Identification Using Oxide Semiconductor Gas Sensor Array and Neural-network Pattern Recognition

In order to identify CH/sub 3/SH, (CH/sub 3/)/sub 3/, C/sub 2/H/sub 5/OH and CO gases in the concentration range of 0.1 to 100 ppm, a gas recognition system using a gas sensor array and neural network pattern recognition has been fabricated. The sensor array consists of such thin film oxide semiconductor sensing materials as 1 wt.% Pd-doped SnO/sub 2/, 6 wt.% AI/sub 2/O/sub 3/-doped ZnO, WO/sub 3/ and ZnO. The principal component analysis and the neural-network pattern recognition analysis were used for the discrimination of gas species and concentrations. Good separation among gases and concentrations was obtained using the principal component analysis. The recognition probability of the neural-network was 100% for each five-trial of twelve gas samples.