Radial basis function artificial neural network model to estimate higher heating value of solid wastes

Abstract

ABSTRACTWaste combustion is considered as a promising method of energy production due to their proper higher heating value (HHV) to be used as fuels. In this study, HHV values are estimated using a radial basis function artificial neural network (RBFANN) combined with Levenberg Marquardt algorithm as the optimization method. HHV values are predicted using mass percentages of water, carbon, oxygen, hydrogen, nitrogen, sulfur, and ash in the solid waste. The proposed RBFANN is then compared with other available correlations, revealing a great accuracy in prediction of HHV values with mean squared error (MSE) and R2 values of 0.248 and 0.997, respectively. Therefore, the proposed RBFANN is considered a reliable predictive tool to estimate the wastes’ HHV values.