Water demand forecasting using the artificial neural network

Abstract

This research presents the application of artificial neural network for forecasting water demand of the customers in the responsible area of the Metropolitan Waterworks Authority. Gross Domestic Product (GDP), Increased Water Tariffs and number of customers are the input of the network that is properly trained with historical data. The result of the forecasting is the water demand. The network used in this study is a two layer feedforward network and the learning process is the backpropagation. This network is trained to be able to forecast water demand accurately. For water demand forecasting of the Fiscal Year 1999-2000, the result from artificial neural network provides more accuracy by having the percentages of error at -0.18% while the result from accrual moving average technique has the percentages of error at 4.06%. Academic Year.