Optimal reservoir rule curves using simulated annealing

Reservoir rule curves are basic guides for the long-term operation of reservoir systems. This paper presents a simulated annealing (SA) algorithm connected with a simulation model to determine optimal reservoir rule curves. The proposed model was applied to the Bhumibol and Sirikit reservoirs in Thailand. The pattern of the obtained rule curves was similar to existing rule curves and the rule curves obtained using genetic algorithms (GAs). The obtained rule curves were used to simulate the Bhumibol and Sirikit reservoir systems with synthetic inflows and these results were compared with the GA-obtained rule curves. The results from the two techniques, in considering both water shortage and excess release of water, were analogous. This indicates that the proposed SA algorithm is effective in determining optimal rule curves for reservoirs.