An intelligent gas sensor application for the discrimination of forane 134a and carbon dioxide gas concentrations: the effect of relative humidity

In the field of smart sensors new applications named electronic noses are now developed and used for environmental air quality control. In our case we are developing such an application based on a tin oxide gas sensor array for the main detection of two gases : a refrigerant gas Forane 134a (Hydro fluorocarbon) and a carbonic gas. Due to the high sensitivity to humidity observed for these types of sensors, we study here the ability of our system to well discriminate the target gases and also their concentrations whatever the relative humidity rate in a wide range. By using a multidimensional pattern recognition method (Discriminant Factorial Analysis) these learning process results are presented and the created decisive laws are used to successfully identify unknown cases.