Improved PSO algorithm based on chaos theory and its application to design flood hydrograph  

作　　者：Si-fang DONG Zeng-chuan DONG Jun-jian MA Kang-ning CHEN 

机构地区：[1]State Key Laboratory of Hydrology-Water Resources and Hydraulic Engineering, Hohai University, Nanjing 210098, P. R. China [2]Department of Remote Sensing Center, China Institute of Water Resources and Hydropower Research, Beijing 100044, P. R. China

出　　处：《Water Science and Engineering》2010年第2期156-165,共10页水科学与水工程（英文版）

基　　金：supported by the National Basic Research Program of China (973 Program) (Grant No.2006CB403402)

摘　　要：The deficiencies of basic particle swarm optimization （bPSO） are its ubiquitous prematurity and its inability to seek the global optimal solution when optimizing complex high-dimensional functions. To overcome such deficiencies, the chaos-PSO （COSPSO） algorithm was established by introducing the chaos optimization mechanism and a global particle stagnation-disturbance strategy into bPSO. In the improved algorithm, chaotic movement was adopted for the particles＇ initial movement trajectories to replace the former stochastic movement, and the chaos factor was used to guide the particles＇ path. When the global particles were stagnant, the disturbance strategy was used to keep the particles in motion. Five benchmark optimizations were introduced to test COSPSO, and they proved that COSPSO can remarkably improve efficiency in optimizing complex functions. Finally, a case study of COSPSO in calculating design flood hydrographs demonstrated the applicability of the improved algorithm.The deficiencies of basic particle swarm optimization （bPSO） are its ubiquitous prematurity and its inability to seek the global optimal solution when optimizing complex high-dimensional functions. To overcome such deficiencies, the chaos-PSO （COSPSO） algorithm was established by introducing the chaos optimization mechanism and a global particle stagnation-disturbance strategy into bPSO. In the improved algorithm, chaotic movement was adopted for the particles＇ initial movement trajectories to replace the former stochastic movement, and the chaos factor was used to guide the particles＇ path. When the global particles were stagnant, the disturbance strategy was used to keep the particles in motion. Five benchmark optimizations were introduced to test COSPSO, and they proved that COSPSO can remarkably improve efficiency in optimizing complex functions. Finally, a case study of COSPSO in calculating design flood hydrographs demonstrated the applicability of the improved algorithm.

关 键 词：particle swarm optimization chaos theory initialization strategy of chaos factor global particle stagnation-disturbance strategy design flood hydrograph 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构] O415.5[自动化与计算机技术—计算机科学与技术]