Artificial neural network model to predict behavior of biogas production curve from mixed lignocellulosic co-substrates

Depletion of fossil fuels and increase in global pollution at an alarming rate has encouraged the researchers to look for environmental friendly and cost effective alternative sources of energy. Biogas production using co-digestion of lignocellulosic biomass with cattle dung is receiving a lot of attention due to its plenty of availability and relatively easy energy conversion technique. Artificial neural network (ANN) is one of the most recent modeling tools, used to solve and predict complicated problems that cannot be explained by conventional methods. The paper demonstrates the modeling and optimization by ANN for prediction of specific biogas production using cattle dung as co-substrates, separately with bamboo dust, sugar-cane bagasse and saw dust in mesophilic as well as in thermophilic condition. In this study, the effect of biogas production parameters such as composition, temperature, and time are considered. Specific biogas production data of 99.7% might be predicted using ANN model within a precision of +/- 10% deviation from the experimental values. The ANN model was used to predict specific biogas production for the substrates at various temperatures. The optimal biogas production was obtained for the mixture of cattle dung and sugar-cane bagasse.