Application of genetic algorithms for irrigation water scheduling

Abstract

The development of a genetic algorithm (GA) to solve an irrigation water scheduling problem is described. The objective is to optimize the utilization of water resources in irrigation systems operating on a rotational basis. An objective function for the water scheduling problem is presented along with constraints that relate to in-field soil moisture balances as well as canal capacities. The approach was applied to a simple and to a more complex test system. Solutions are presented using a GA in different formulations and comparisons made between these. Results demonstrate that GAs are capable of solving water scheduling problems, including those with water stress. In water stress conditions the GA approach can provide uniformity in soil moisture content in schemes within a system if formulated with a 0-1 approach. An application to the Pugal branch canal in the Indira Gandhi Nahar Pariyojana (IGNP) irrigation system in north-west India has demonstrated that the approach is robust and can produce appropriate schedules under extreme conditions of water stress. The GA approach is a useful tool for water scheduling in complex systems. Copyright © 2004 John Wiley & Sons, Ltd.