Pattern recognition of gas sensor array using characteristics of impedance

In this paper, impedance variations of SnO2-based gas sensor array are meas ured in gases. Back propagation algorithm of pattern recognition [1,2] is u sed to distinguish a species and concentration of gases with analog signals of impedance variation from sensor array. When the impedance variation pre sents as the quantities of rising and falling time to 100 kHz pulse input s ignal of a prepared circuit in gases, the sensor array system distinguishes gas species and concentration with back propagation algorithm which deals with rising time and falling time from sensor array. SnO2-based sensing mat erials on a sensor array are SnO2-Pt, SnO2-La2O3, SnO2-Au/Pd, and SnO2-Pd-V 2O5 using thick film process techniques. The proposed sensor array for patt ern recognition with impedance variation is proper to enhance stability, se lectivity, and reproducibility to gases. (C) 2001 Elsevier Science B.V. All rights reserved.