Determination of influence parameters on concentration of carbon monoxide in exhaust gases by using artificial neural networks (ANN)

Abstract

Constant tightening of legal norms requires constant research activities relating to the reduction of exhaust emissions. So far, experiments were almost always used to determine the concentration of a harmful product in the exhaust gases. Thanks to progress in technology and computational intelligence, it is possible to approach this problem in a more elegant way. By developing an artificial neural network (ANN) model, it is possible to predict concentration of some harmful products in exhaust gases with great accuracy. It is proven that, with ANN, something like this is possible by using an appropriate training tool, even for a very small number of baseline data.