Multi-reservoir System Operation in Drought Periods with Balancing Multiple Groups of Objectives

Abstract

Water resources systems should be operated to balance multiple conflicting objectives accounting for a variety of services. In this field, the evolutionary algorithms are very helpful since complex simulation models can be directly embedded within them, and they are also powerful for deriving the trade-off among conflicting objectives in multi-objective optimization problems. In this study, the WEAP (Water Evaluation And Planning) water resources simulation model, and the NSGA-II (Non-dominated Sorting Genetic Algorithm) multi-objective optimization model, are employed and coupled to extract optimal trade-off among intra-basin, interbasin, hydropower and environmental flow objectives in a Multi-reservoir system. In such a context, there is not a single operating policy that optimizes simultaneously all the water purposes. Therefore, opposite views and disputes on the selection of the best policy among different Pareto solutions arise. In this regard, Game Theory, the formal study of mathematical models of conflict and cooperation between intelligent, rational decision-makers, is adopted to decide on the best hedging rules and operating rule curves among Pareto alternatives. The Quantal Response Equilibrium (QRE) which is a solution concept in Game Theory is used here. To investigate the models, Zohreh three-reservoir multi-purpose system in the southwestern Iran with four different objectives is studied.