Particle Swarm Optimization with Diversity-Controlled Acceleration Coefficients
Jing Jie, Jianchao Zeng · 2007
In order to overcome the premature convergence, the paper introduced a negative feedback mechanism into particle swarm optimization and developed an adaptive PSO. The improved method takes advantage of the swarm-diversity to control the tuning of the acceleration coefficients (PSO-DCAC). Through the feedback control of the diversity, PSO-DCAC can manipulate the weight of the cognitive part and the social part to fluctuate with the search state, which in turn can adjust the exploration and exploitation adaptively and contributes to a successful global search. The proposed PSO-DCAC was applied to some well-known benchmarks and compared with the other notable improved PSO. Experimental results show diversity-controlled acceleration coefficients is a feasible technique to improve the global performance of PSO and performs very well on the complex optimization problems.