Optimal Irrigation Scheduling with Evolutionary Algorithms

摘要

Efficient irrigation is becoming a necessity in order to cope with the aggravating water shortage while simultaneously securing the increasing world population's food supply. In this paper, we compare five Evolutionary Algorithms (real valued Genetic Algorithm, Particle Swarm Optimization, Differential Evolution, and two Evolution Strategy-based Algorithms) on the problem of optimal deficit irrigation. We also introduce three different constraint handling strategies that deal with the constraints which arise from the limited amount of irrigation water. We show that Differential Evolution and Particle Swarm Optimization are able to optimize irrigation schedules achieving results which are extremely close to the theoretical optimum.