Ant Colony Optimization Applied to Water Distribution System Design: Comparative Study of Five Algorithms

Water distribution systems (WDSs) are costly infrastructure, and much attention has been given to the application of optimization methods to minimize design costs. In previous studies, ant colony optimization (ACO) has been found to perform extremely competitively for WDS optimization. In this paper, five ACO algorithms are tested: one basic algorithm (ant system) and four more advanced algorithms [ant colony system, elitist ant system, elitist-rank ant system (ASrank), and max-min ant system (MMAS)]. Experiments are carried out to determine their performance on four WDS case studies, three of which have been considered widely in the literature. The findings of the study show that some ACO algorithms are very successful for WDS design, as two of the ACO algorithms (MMAS and ASrank) outperform all other algorithms applied to these case studies in the literature.