HIGH-PRECISION NEURAL PREPROCESSING FOR SIGNAL ANALYSIS OF A SENSOR ARRAY

An array of semiconducting thin-film sensors is used for high-accuracy determination of gas concentrations in a binary mixture of H2S and NO 2. For this purpose we developed a newly conceived pre-processing meth od for data feeding of a standard neural network. Pre-processing explo its dimension compaction of the vectorial signal of the sensor array. This operation is being accomplished by an auxiliary auto-associative network, separately trained from the main one. Our results present unp recedented accuracy even when applied to a validation-data set and app ear to be of widespread application. (C) 1998 Published by Elsevier Sc ience S.A. All rights reserved.