LAUSR

In this paper, a catfish-effect multi-objective particle swarm optimization algorithm (CEMOPSO) is proposed for optimizing the coordinative flow-sediment scheduling in a reservoir. In the proposed CE-MOPSO, the driven effect of catfish particles is introduced to improve the convergence and diversity of solutions. The performance of the proposed CE-MOPSO was verified using a classical biobjective test function (ZDT3), and it was found that compared with MOPSO and Sigma-MOPSO algorithms, the CE-MOPSO showed better convergence to the true Pareto optimal fronts, and provided better diversity and uniformity for the Pareto frons with smaller values of convergence index and diversity index. After the successful validations in simulation studies, the proposed approach was then applied to a real case study in the Three Gorges Reservoir in China. Our results showed that the obtained Pareto solution set effectively approximated the true Pareto optimal frontier during the process of evolution. The scheduling results of CE-MOPSO revealed the relationship between power generation and sediment deposition in ten years and can be used to develop reservoir operation policies and plan sediment trapping and flow operations in real time. These results suggest that the proposed CE-MOPSO approach is efficient and effective in managing multi-objective water resources and hydrologic problems.