ANN-based Determination of Optimum Working Conditions of Residential Combustors with Respect to Optimum Insulation

In this study, optimum working conditions of combustion chambers were determined taking optimum insulation thickness into account. With this aim, different fuels, such as Tuncbilek lignite and natural gas, were selected to be used in the combustion chamber. The combustion process was then evaluated using exergo-economic analysis. The results obtained from this analytic evaluation were used to train a multilayer artificial neural network model with the back-propagation learning algorithm with three different variants, namely, Levenberg-Marguardt, Pola-Ribiere conjugate gradient, and scaled conjugate gradient. The most suitable algorithm was found to be Levenberg-Marguardt with eight neurons in a single hidden layer.