Extension of Parametric Rule with the Hedging Rule for Managing Multireservoir System during Droughts

Abstract

In contrast to most common methods used in optimal control of reservoir systems requiring a large number of decision variables, parametric rule can make a radical reduction in the number of decision variables without yielding inferior solutions. However, parametric rule employs the standard operating policy to determine releases of reservoirs as much as demand only if there is enough water in the system, which may result in single periods of severe short supply during droughts. The purpose of this paper is to devise an operating rule for multireservoir system by combining parametric rule with the hedging rule to avoid catastrophic water shortage during droughts. In this way, decision variables to be optimized not only make a significant reduction compared with traditional operating rules, but also severe short supply during droughts can be controlled effectively. This paper employs a water supply multireservoir system in northern China to explore the changes of shortage characteristics produced by the proposed rule over a long horizon. In the case study, particle swarm optimization algorithms with a simulation model are used to optimize the decision variables. The results indicate that the extended parametric rule has a significant advantage over the classic parametric rule in dealing with the multireservoir operation problem during droughts.