Hierarcical strategy for quantification of NOx in a varying background of typical exhaust gases

Silicon carbide based metal insulator field effect transistors (MISiC FET) with a catalytic gate metallization were used in temperature cycled operating mode (TCO) in order to improve the selectivity of the sensor. This approach obtaining multiple data from a single sensor, therefore known as a virtual multisensor, was originally developed for metal oxide sensors but earlier work proved the suitability for MISiC FETs as well. This strategy was now tailored to the quantification of NO_x in a mixture of typical exhaust gases (CO, HC, plus NH₃). Data was evaluated with multivariate statistics e.g. Linear Discriminant Analysis. In order to suppress the influence of varying background a hierarchical approach was used. Results show that quantification of NO_x is possible even in a changing background.