摘要
针对困扰粒子群算法"早熟"和收敛慢的2大难题,提出了一种跟随变异粒子扰动变化的惯性权重策略,将粒子群体一致性变化趋势与粒子变异操作相关联,以此来增加粒子的多样性;根据粒子变异程度的大小对惯性权重作出相应调整,使粒子在提高全局寻优能力的同时,又很好地改善了收敛精度和收敛速度,避免因惯性权重单边非线性变化而导致粒子群全局寻优能力稳定性不佳的问题。仿真对比表明,改进的算法较好地避开了局部最优解的干扰问题,具有收敛速度快、寻优精度高等优势。
Aiming at the problems of the particle swarm algorithm ＂premature＂and slow convergence,the inertia weight adjustment strategy with the mutation particles disturbance change was presented. In order to increase the diversity of the particles,the consistent of the particles group changing tendency was related with the particle mutation. The inertia of the weight of the particles was put forward with a corresponding adjustment according to the degree of variation the mutation particles,and the global searching ability of the particles is not only improved,but also that the convergence precision and speed of the particles are advanced,the poor stability problem of the particles swarm optimization ability,which is generated due to the inertia weight unilateral nonlinear change,is avoided. Simulation results show that the improved algorithm has the advantages of fast convergence speed,high precision optimization and so on.
出处
《四川兵工学报》
CAS
2015年第1期106-110,共5页
Journal of Sichuan Ordnance
基金
军队研究生课题(2011JY002-422)的资助
关键词
变异粒子
惯性权重
粒子群算法
mutation particle
inertia weight
particle swarm optimization
