Estimation of Biogas Production Rate in a Thermophilic UASB Reactor Using Artificial Neural Networks

Abstract

Biogas production rate was modeled and estimated in a thermophilic upflow anaerobic sludge blanket digester. Data set covers a time period of both steady-state conditions and an abnormal operation condition, i.e., organic loading shocks. Multilayer neural networks topology was used as the modeling tool. Half of the experimental data were used for the training of the model and the remaining half were used for the testing stage. Model results were evaluated from the point of view of both steady conditions and abnormal conditions. It was seen from the time series trends of the estimated data that biogas production rates at steady state operation conditions were closely estimated by the model while the results for organic loading shocks were sufficiently followed. Artificial neural network models gave encouraging estimation results for the online control of thermophilic reactors.