Coupled-operations model and a conditional differential evolution algorithm for improving reservoir management

Abstract

This paper proposes a flexible methodology for finding the optimal control of reservoir operations, which is adopted for simulating a storage system including a large reservoir simulation model and a numerical search method for searching decision variables. The optimization model used was a conditional differential evolution algorithm (CDE). The model was connected to reservoir simulation model for searching the decision variables in reservoir management. The reservoir management model was applied to Lampao reservoir in Kalasin province located in the northeast of Thailand. This study considers the criteria of minimum water shortages and downstream flood control. There were synthetic inflow data for evaluating the performance of the reservoir operating rule curves. It was found that the reservoir operating rule curves from the CDE showed a lower frequency and quantity of water shortage than the existing reservoir operating rule curve. In addition, the flood frequency and the flood quantity were reduced. The results also demonstrated that the CDE provided better reservoir operation in any probability of inflow situations. Further, it indicated that the reservoir operating rule curves obtained from this research could effectively support the increasing water requirements to a growing population. Key words: A conditional differential evolution algorithm (CDE), simulation model, optimization technique, reservoir rule curves