Artificial neural network based modeling of biomass gasification in fixed bed downdraft gasifiers

The study attempts at developing an artificial neural network (ANN) based model of biomass gasification in fixed bed downdraft gasifiers. The study is a novel attempt in developing an ANN based model of biomass gasification in fixed bed downdraft gasifiers as there are very few reported studies of ANN based modeling of biomass gasification in general and even fewer in the field of fixed bed downdraft gasifiers. In fact, downdraft gasifiers are one of the most widely used type of gasifiers for small scale operation. The ANN based models were formulated to predict the product gas composition in terms of concentration of four major gas species viz. CH4%, CO%, CO2% and H-2%. The input parameters used in the models were C, H, 0 content, ash content, moisture content, and reduction zone temperature. The architecture of the models consisted of one input, one hidden and one output layer. Reported experimental data were used to train the ANNs. The output of the ANN models were found to be in agreement with experimental data with an absolute fraction of variance (R-2) higher than 0.99 in the cases of CH4 and CO models and higher than 0.98 in the case of CO2 and H-2 model. The results show the possibility of utilization of the model to predict the percentage composition of four major product gas species (CH4, CO, CO2 and H-2). The relative importance of the input variables was also analysed using the Garson's equation. (C) 2017 Elsevier Ltd. All rights reserved.