Abstract:
Water resource management systems are indispensible tools for provision of adequate water services across towns, municipalities and cities. This study attempts to develop and compare three variants of artificial neural networks to predict monthly water demand using socio-economic and demographic data of Oman. The results showed that the cascade forward neural network outperformed Elman and feed forward neural networks. It is envisaged that these water demand foresting models would provide useful and relevant information to top managers to make operational, tactical and strategic decisions to meet the demand for water supply efficiently.
|