SIMULTANEOUS QUANTIFICATION OF CARBON-MONOXIDE AND METHANE IN HUMID AIR USING A SENSOR ARRAY AND AN ARTIFICIAL NEURAL-NETWORK

The simultaneous quantification of carbon monoxide and methane in humi d air is presented. The response of a three-sensor array, including an undoped and a platinum doped tin dioxide sensor showing non-ideal sel ectivity and a humidity sensor is fed into an optimised feed forward b ack propagation artificial neural network in order to obtain the carbo n monoxide and methane concentration as network output. The gaseous en vironments under study were ternary mixtures in the concentration rang es of 0-0.5% methane, 0-1000 ppm carbon monoxide and 0-60% relative hu midity at 20 degrees C. The network structure, network output with res pect to a priori known test concentrations and the influence of the si ze of the training data set is discussed. (C) 1997 Elsevier Science S. A.