Performance evaluation and optimization of irrigation canal systems using genetic algorithm

Improving the performance of irrigated agriculture is essential for increasing productivity to feed the rapidly growing world population. Methodological difficulties exist in performance evaluation and improvement of irrigation systems due to the complexity of the problem and the lack of analytical tools to evaluate performance, to establish performance standards, and to identify the significance of the various factors which contribute to performance improvement. To resolve these methodological difficulties, a framework for performance evaluation and improvement of irrigation systems is developed. A system approach is emphasized. The irrigation system manager is responsible to satisfy the internal and external expectations from the irrigation system in an optimum fashion. He has control over the variables affecting the operation of the irrigation system. Therefore, his perspective is selected as the ideal framework. Appropriate performance indicators based on different water delivery methods are selected and a simulation technique is used for their analysis. The performance evaluation model is applied to a canal system typical of the Eastern Irrigation District in southern Alberta, Canada. The model is applied to find operational alternatives which will improve performance under conditions of uncontrolled discharge fluctuation at the canal intake. The genetic algorithm is selected for use in an irrigation system optimization model because of its simplicity, flexibility, and robustness. The Performance Optimization Model (POM) is developed and linked with Irrigation Conveyance System Simulation (ICSS) model to identify the optimum improvement alternatives. The POM-ICSS model employs a system approach and solves multi-variable, multi-objective, and multi-constraint problems related to irrigation system performance. The model provides optimum operational instructions, realistic performance standards, and indicates the contribution of the different factors to the performance of irrigation systems. The model is applied on a canal system for both steady and unsteady flow situation. Optimum performance for a variety of physical and management alternative is determined. The suggested methodology and the POM-ICSS model demonstrate that the system analysis approach is a powerful decision support facility for the irrigation system manager, as well as all of the stake holders, for improvement of the performance of irrigation systems world wide.