Determination of optimal water resource management through a fuzzy multiobjective programming and genetic algorithm: Case study in Kinman, Taiwan

Given the increase in demands on water resources, optimal management policy has been sought for minimum resource utilities and maximum environmental benefits. Since the water resource system is complicated and has many uncertainties, artificial intelligence techniques, like fuzzy set theory or a genetic algorithm (GA), enable us to identify optimal decision variables more efficiently and incorporate uncertainty considerations. In this study, multiobjective programming was developed for a realistic application in Kinmen Island, Taiwan, against the scarce water resource. Three alternatives, including a desalination facility, additional activated carbon equipment for the current wastewater treatment plant (WTP), and constructed wetlands were designed to improve water quality and to ensure sufficient water supply. The GA technology is employed to search for the best solutions from the three alternatives. In addition, the fuzzy goal of cost and water quantity represented is utilized to determine the optimal control policy by the degree of satisfaction. In the optimal management of this case study, the desalination facility receives 17.8% of the treatment water, and 20.7 and 61.5% of the treatment water allocated to the current WTP and constructed wetland, respectively. The GA and fuzzy multiobjective programming are proven to facilitate the determination of the optimal solution for Kinmen's water resource management.