MOPSO in Multipurpose Operation of Single-Reservoir System

Complexity of engineering problems, especially in the field of water resources, makes traditional optimization tools less efficient and enhances the popularity of evolutionary algorithms. Traditional optimization methods present just a single point as the optimal solution even in multi-objective problems. Thus, the main necessity to apply evolutionary algorithms is to present a set of nondominated solutions/Pareto-fronts as the optimal/near-optimal solutions, whereas most of problems in the field of water resources management are considered as a compromise between several objectives. In this paper, the Multi-Objective Particle Swarm Optimization (MOPSO) is applied to the operation of a multipurpose reservoir system. The objectives are: (1) maximization of reliability; (2) minimization of vulnerability; and (3) minimization of resiliency of the reservoir system, satisfying the downstream demand. The proposed algorithm is compared to a traditional method (weighting method). Results show that the MOPSO algorithm is successful in finding near-optimal, Pareto-front solutions.