Optimization Of Fuzzy Programming For Reservoir Operation

The framework of this study includes a fuzzy programming and three optimization algorithms for reservoir operation. The fuzzy programming theorem is adopted to construct a suitable set of objective function and constraints of reservoir operation, while the optimization methods are used for searching long-term operation histograms. The optimization methods include the genetic algorithm (GA), the simulated annealing (SA), and a hybrid module, the annealing genetic algorithm (AG). The Shi-Man reservoir, Taiwan, is used as a case study. Its observed monthly inflow data in three different hydrological conditions are implemented to investigate the model performances through different optimization approaches. The degree of satisfactory, Generalized Shortage Index (GSI), and total shortage amount are the criteria for evaluating the model­Ýs performance. For the purpose of comparison, the current used M-5 operation rule is also performed. The results demonstrate that all three optimization algorithms (i.e. GA, SA, and AG) do come out better performances, in terms of larger satisfactory and smaller GSI and total shortage amount, than the current used M-5 operation rule in all the cases, while the annealing genetic algorithm has the greatest capability and highest efficiency for optimizing the fuzzy programming model in reservoir operation.