S Bahrami - Department of Chemical Engineering, Faculty of Chemical Engineering & Petroleum, University of Tabriz, 5166616471, Iran
A Farzi - Department of Chemical Engineering, Faculty of Chemical Engineering & Petroleum, University of Tabriz, 5166616471, Iran
H.R Khaledian - Department of Chemical Engineering, Faculty of Chemical Engineering & Petroleum, University of Tabriz, 5166616471, Iran
A Tarjomannejad - Department of Chemical Engineering, Faculty of Chemical Engineering & Petroleum, University of Tabriz, 5166616471, Iran

Artificial neural networks are a suitable algorithm for modeling of chemical processes .In this study the effective factors of catalysts and also the reaction temperature were used in NO+CO reduction over LaCo0.5B0.5O3 (B=Cr, Cu, Mn) perovskite-type nanocatalysts to introduce a neural network. The network was made of 3 layers. One input layer, one hidden layer and an output layer and its training function was Levenberg-Marquardt. Also the optimum number of neurons in hidden layer was19. The correlation coefficient R2 for all data was equal to 0.9967, which means the values which were obtained from the discussed ANN were in a good agreement with the experimental data

modeling, NOx reduction, ANN