Application of simulated annealing (SA) techniques for optimal water distribution in irrigation canals

One of the contributing factors to the poor performance of irrigation systems is improper water distribution. Traditionally, in irrigation canals water distribution schedules are determined based on descriptive methods. Application of optimization methods could improve performance of water delivery. In a case study, a single‐objective zero–one programming method was used to develop an optimal water delivery schedule. Considering several objectives of irrigation systems and the difficulties of zero–one programming, a multi‐objective simulated annealing optimization technique is utilized. Based on water requirements at delivery points, the developed model determines time and rate of water delivery. The canal capacity and operational stages are minimized and water requirements are satisfied as closely as possible. In order to test the applicability of the proposed model, the model is applied on the Feng‐Jia‐Shan Canal in China and compared with the results of Wang. The results show that the multi‐objective optimization approach resulted in lower operational stages, while the canal capacity is reduced to the same level as Wang's study. The model is applied on the AMX canal of the Varamin irrigation network in Tehran, Iran. The model benefits from the advantages of SA techniques and could take into account several objectives and avoid the difficulties of zero–one programming. This model could be used as an efficient tool for planning water delivery schedules in irrigation canals. Copyright © 2005 John Wiley & Sons, Ltd. Un des facteurs de contribution à la basse exécution des systèmes d'irrigation est la distribution inexacte de l'eau. Traditionnellement, dans la distribution de l'eau de canaux d'irrigation des programmes sont déterminés basés sur des méthodes descriptives. L'application des méthodes d'optimisation a pu améliorer l'exécution de la livraison de l'eau. Dans une étude de cas, une méthode de programmation de zero–one simple‐objectif a été employée pour développer un programme optimal de la livraison de l'eau. Vu plusieurs objectifs des systèmes d'irrigation et des difficultés de zero–one programmation, la technique de recuit simulée (SA) multi‐objective d'optimisation est utilisée. Basé sur des conditions de l'eau aux points de livraison, le modèle développé détermine la période et le taux de la livraison de l'eau. La capacité de canal et les étapes opérationnelles sont réduites au minimum et des conditions de l'eau sont satisfaites aussi étroiteme