基于群体多样性反馈控制自组织微粒群算法_介婧.pdfVIP

ISSN 1000- 1239PCN 11- 1777PTP Journal of Computer Research and De elopment 45( 3) : 464~ 47 1, 2008 1, 2 2 1 介 婧 曾建潮 韩崇昭 1( 710049) 2 ( 030024) (jjing277@ sohu1 com) Self-Organized Particle Swarm Optimization Based on Feedback Control of Diver- sity 1, 2 2 1 Jie Jing , Zeng Jianchao , and Han Chongzhao 1( School of Electronic and Inf ormation Engineering , X i. an Jiaotong University, X i. an 7 10049) 2 (Div ision of System Simulation Comp uter App lication, aiyuan University of Science echnology, aiyuan 030024) Abstract Particle sw arm optimization ( PSO) is a no el sw arm intelligence algorithm inspired by certain social beha ior of bird flocking originally1 Since proposed in 1995, the algorithm pro ed to be a alid opti- mization technique and has been applied in many areas successfully1 How e er, like others e olutionary algo- rithms, PSO also suffers from the premature con ergence problem, especially for the large scale and com- plex problems1 In order to alle iate the premature con ergence problem, the paper de elops a self-organized PSO( SOPSO) 1 SOPSO regards the sw arm as a self-organized system, and introduces negati e feedback mechanism to imitate the information interaction betw een the particles and the swarm background1 Consid- ering sw arm di ersity is a key factor influencing the global performances of PSO, SOPSO adopts swarm di- ersity as main dynamic information to control the tuning of parameters through feedback, w hich in turn can modify the particles to di erge or con erge adapti ely and contribute to a successful global search1 The proposed methods are applied to some complex function optimizations and compar