Prediction of Water Demands in a Water Treatment Plant Using an Artificial Neural Network Model
To provide improvements in efficiency and the ability to respond to changing external conditions, an artificial neural network (ANN) model is used to characterize the contents of reservoir(s) and water tower(s) sufficient to meet water demands. Maintaining water levels and daily treatment quantities can be effective to reduce the formation of disinfection byproducts (DBPs). Predictive models are developed to investigate the effects of maximum daily temperatures, incoming solar radiation and total daily precipitation (which influences water demands) and DBP formation at the water treatment plant.
ANNs are semi-parametric regression estimators, and are well suited for predicting water demands and water quality as they can approximate virtually any function, to varying degrees of accuracy. In this application, predictions of water demand and DBPs are obtained using a simple back-propagation neural network. A comparative evaluation of the stepwise regression method and the ANN model are provided. The results show that the ANN obtains better accuracy than the regression method: it produced R2 = 0.84 as compared to R2 = 0.71 obtained by standard regression.
This paper is only available in PDF Format:
View full text PDF