基于鲶鱼效应多目标粒子群算法的水库水沙联合优化调度  

The Optimal Dispatch of Water and Sediment in Reservoir Based on the Catfish Effect Multi-objective Particle Swarm Algorithm

作　　者：刘方[1] 纪昌明[1] 向腾飞[1] 周茜[1] 

机构地区：[1]华北电力大学可再生能源学院,北京102206

出　　处：《中国农村水利水电》2012年第11期4-8,共5页China Rural Water and Hydropower

基　　金：国家自然科学面上基金资助项目(51179069);国家自然科学基金资助项目(40971300)

摘　　要：针对水库水沙联合优化调度多目标、高维、非线性和难以求解的特点,将鲶鱼效应机制引入多目标粒子群算法中提出基于鲶鱼效应的多目标粒子群算法,该算法在利用收敛速度较快的Sigma方法的基础上,通过触发鲶鱼启发器引入外部鲶鱼粒子,利用鲶鱼粒子对种群的驱赶效应增加种群多样性,从而提高算法的收敛性和非劣解集的多样性;数值分析证明,与MOPSO和σ-MOPSO相比,该算法的效率和质量更高,同时三峡水库实际算例也表明,该算法能给出代表整个可行调度空间、收敛较好、分布均匀的Pareto最优前沿,具有较好的适应性。In the light of the characteristics of the water-sediment coordinative optimized dispatch, such as multi-objective, multi-di- mension, non-linearity and difficulty in being solved, the catfish effect mechanism is introduced into the multi-objective particle swarm optimization called catfish effect multi-objective particle swarm optimization. On the basis of the advantage of the fast convergent of Sigma Method, catfish particles are produced by the catfish generator in the evolution that will trigger the driven influence of catfish effect to improve the convergence and diversity of solutions. The numerical analysis has addressed that, compared with the MOPSO and σ-MOPSO algorithm, the improved algorithm has a higher efficiency and quality. And the case study also shows that, the algo-rithm has a high degree of adaptability because it can generate the final Pareto frontier that represents the whole feasible scheduling space and has a high degree of convergence and diversity.

关 键 词：水沙联合调度 多目标粒子群算法 鲶鱼效应 Sigma方法 

分 类 号：TV12[水利工程—水文学及水资源]