An intelligent gas sensing system

Abstract

Electrically conducting organic polymers are widely used as a means of gas, odour or aroma analysis using multi-element array techniques coupled with direct current (d.c.) interrogation techniques. Recently it has been established that the use of alternating current (a.c.) interrogation gives rise to improved performance. In addition, the need to use multi-element arrays is much reduced since a single sensor can be interrogated at a wide range of frequencies. This gives rise to much increased information content for the measurements. This paper describes the use of alternating current (a.c.) interrogated conducting organic polymers coupled with neural network pattern recognition techniques for a system to determine the compositional fraction of volatile vapour mixtures. Experiments have been conducted on binary, tertiary and quaternary mixtures of vapours and compositional fractions have been determined to within 5%.