Thermally pulsed metal oxide gas sensor combined with a colorimetric gas sensor for the detection of trace gases

A low-cost gas sensor array with low-power consumption consisting of two chemical sensors based on different principles and a smart pattern recognition technique is presented. The sensor array consists of a thermally pulsed metal oxide gas sensor combined with an optical sensor based on the colorimetric principle. Metal oxide gas sensors suffer from inadequate selectivity and colorimetric sensors from slow response times. The aim of the presented work is to compensate these drawbacks through intelligent data fusion using an artificial neural network. It can be shown that the disadvantages of the two types of sensors can be compensated by the data fusion. The proof of principle is demonstrated by trace gas measurements with carbon monoxide, nitrogen dioxide and ammonia.