Ant Colony Optimization for the Design of Small-Scale Irrigation Systems

The optimal design of sprinkler irrigation systems is a complicated nonlinear programming problem that is related to the performance of the system and meanwhile an economic problem to farmers in developing countries. Ant colony optimization (ACO), a meta-heuristic algorithm with the strategies inspired by foraging ants, was considered. Exactly an Ant Cycle System was proposed to solve this problem. The performance of ACO was compared to that of Genetic Algorithm (GA), and the optimal results were further validated by field tests on four small-scale irrigation systems. In the optimization model, the objective function was minimizing the specific energy consumption subject to the constraints of pipe diameters, number of sprinklers and working pressure of the end sprinkler along the pipeline and pump-pipeline cooperation conditions. In the design of ACO, head loss between adjacent sprinklers was introduced in the heuristic function to represent the distance between two cities in a Travelling Salesman Problem (TSP). And the fitness composed of the specific energy consumption dealt with penalty function was taken instead of the total length of a route in the pheromone updating. The results indicate that the specific energy consumption has been decreased in average by 12.45 % through ACO, 10.27 % through GA and 11.27 % from field tests compared to that in the initial configurations with irrigation uniformities higher than 75 % in the field tests. ACO implementation outperforms genetic algorithm in efficiency and reliability especially in larger systems. The ACO may provide a promising approach for the optimization of irrigation systems.