MODELING OF LAB-SCALE ACTIVATED SLUDGE REACTOR USING ARTIFICIAL NEURAL NEWORKS
Keywords:
Artificial Intelligence (AI), Artificial Neural networks (ANN), Activated sludge process (ASP), non-linear auto regressive with external inputs (NARX), wastewater treatment plant (WWTP)Abstract
Artificial Intelligence (AI) models are being used for the simulation and control of biological processes in
wastewater treatment plant (WWTP). These models can be described as computational methodologies which reflect the
behavior of non-linear relationships between cause and effects irrespective to the process. In this study, artificial neural
network (ANN) models were used as an AI method for simulation and prediction of effluent parameters in Activated
sludge process (ASP). The effluent COD as a model output was predicted by taking time varying input parameters such
as pH, TDSinf, BODinf, CODinf of daily data from the measured parameters of ASP. The model was developed by
using artificial neural network for multistep- ahead prediction with non-linear auto regressive with external inputs
(NARX) tool in MATLAB/Simulink(R2012a). The script was written in the MATLAB with training, validation and testing
as the stages of prediction. From the analysis of the results obtained by this model, it was found that the value of
regression coefficients for the best fit model was 0.8095 with hidden layer size-8 and trainlm as training function.
