Application of ANFIS, ARIMA and hybrid models in water demand forecasting

Abstract

Water demand is the total amount of water required for ecosystem functions and processes and out-of-stream uses. Water demand is important for domestic use, residential, irrigation, and industrial activities. Hence, water demand forecasting has become an essential component in effective water resources planning and management. In this study, water demand data is obtained from Syarikat Air Johor (SAJ) ranging from 1995 until 2011. Data from two areas of Johor that are Muar and Batu Pahat were chosen for analysis. For this study, ANFIS, ARIMA and hybrid models are three methods used for this water demand forecasting. The effectiveness of these models has been investigated based on performances of MAE and MSE. It can be concluded that the hybrid model is an effective method to forecast water demands for Muar and Batu Pahat. The results of the hybrid model show that hybrid model can be applied successfully to establish time series forecasting models, which could provide accurate forecasting and modelling time series