Wastewater treatment plant performance evaluation using artificial neural network model

Abstract

The final protection of wastewater pollutants to the environment is wastewater treatment plant (WWTP). This Research paper intends to develop a model from the past operational data of WWTP using artificial neural network (ANN) for the evaluation of the plant performance. The model was developed using four years’ operational data of Bandar Tun Razak sewerage treatment plant (STP) Kuala Lumpur, Malaysia. Matlab version 2008b software was used in training the ANN model. The model shows a high prediction capacity for multiple parallel outputs of effluent biochemical oxygen demand (BODe), chemical oxygen demand (CODe), suspended solid (SSe) and Ammonical nitrogen (NH3-Ne) than single output of each of the parameters individually. The coefficient of determination (R) and mean square error of the best model were 0.9449 and 2.86 respectively.