CAJ | 학술논문

对于粒子群优化算法易陷入局部最优的缺陷，反向学习策略对其的改进取得了较好的效果。然而，反向学习策略需要结合其它策略来提高算法后期的全局搜索能力，针对此缺陷，根据光的折射原理对反向学习策略的反向过程进行改进，提出反向学习的统一算法模型及基于折射原理反向学习模型的改进粒子群算法。实验与分析表明，与其它基于反向学习的粒子群算法相比，该模型更有效地改进了所提算法的全局搜索能力，提高了种群的多样性，从而提高了算法的收敛速度以及优化精度。
대우입자군우화산법역함입국부최우적결함，반향학습책략대기적개진취득료교호적효과。연이，반향학습책략수요결합기타책략래제고산법후기적전국수색능력，침대차결함，근거광적절사원리대반향학습책략적반향과정진행개진，제출반향학습적통일산법모형급기우절사원리반향학습모형적개진입자군산법。실험여분석표명，여기타기우반향학습적입자군산법상비，해모형경유효지개진료소제산법적전국수색능력，제고료충군적다양성，종이제고료산법적수렴속도이급우화정도。
One of shortcomings found in the particle swarm optimization algorithm is that it is easy to fall into local opti-mum,and the opposite learning strategy has a good effect on the improvement of this shortcoming.However,to improve the global search ability by using the opposite learning strategy it is necessary that in the late algorithm other strategies are combined to oppo-site learning strategy.To overcome this shortcoming,this paper improves the opposite process of the opposite learning strategy ac-cording to the refraction principle of light,and proposes the unified model of opposite-based learning(UOBL)and the improved par-ticle swarm optimization algorithm based on the opposite learning model of the principle of refraction(refrPSO).Experiment results and analysis show that the model improves the global search ability of the refrPSO algorithm more effectively compared with other particle swarm algorithm based on opposite learning and the diversity of the population.Because of these improvements,the refrPSO enhances the convergence speed and the accuracy of optimization.