Artificial Neural Network Modeling for Predicting Organic Matter in a Full-Scale Up-Flow Anaerobic Sludge Blanket (UASB) Reactor

The aim of this study is to propose a method for constructing Artificial Neural Network (ANN) models and evaluating their performance based on the application of two methods for the selection of the ANN topology: the dynamic division method (cross-validation or dynamics-split) (DDM) and the static-split method (SSM). The two methods are compared and applied to predict the amount of organic matter in an up-flow anaerobic sludge blanket (UASB) reactor operated at full scale. The performance of the ANN models was assessed through the coefficient of multiple determination (R-2), the adjusted coefficient of multiple determination (R-adj(2)), and the root mean square error (RMSE). The comparison reveals that the DDM accurately selects the best model and reliably assesses its quality.