Review of hybrid evolutionary algorithms for optimizing a reservoir

Abstract

Many studies present hybrid algorithms to solve multiobjective uses of a reservoir. The reservoirs presented in this review are used for flood control, hydropower generation, ecological flow requirement and water distribution systems. While flood control and hydropower are main function of a reservoir, ecological flow requirement are shown to be an important part that should be incorporated into reservoir operation models. Evolutionary algorithms are shown to be capable of solving complex reservoir operation models with fast convergence rate. A review of different algorithms in solving different reservoir operation problems is presented. The results generated by these algorithms are effective, competitive, comparable and applicable. The algorithms present the solutions to the computationally expensive models in an efficient way. Systematic ways of solving different reservoir operation models are presented. Many models reviewed involve reservoirs operated in single objective, multiobjectives, single reservoir and multireservoir. Real time operation is shown to be superior to normal operation. The results generated by the evolutionary algorithms presented show that the algorithms are capable of solving complex and multidimensional problem of water resources. The non-dominated solutions generated are many and spread widely on the Pareto-optimal front.