IDENTIFICATION OF COMBUSTIBLE GASES WITH CATALYTIC SENSORS

Thermocatalytic gas sensors are among the most widely used for measuring concentrations of combustible gases. However, in spite of the variety of advantages of this type of sensors, there are also shortcomings, one of which is low selectivity to target combustible gases. This study presents the results of the application of machine learning methods for the identification of flammable gases in the air using thermocatalytic sensors. The sensor signal measurements were performed using the multistage measuring pulse method, which was described in our previous papers. To enrich the sensor signal with information about ambient gas composition, the number of stages of the method was drastically increased. Methane and vapors of acetone, ethanol and gasoline were used as the target gases. The multivariate data was processed with the support vector machine method. The results show that the proposed method allows for gas identification using a single catalytic gas sensor. The method can be used in the development of highly selective gas detectors responding only to target gases.