摘要
目前机电产品系统的优化设计过程多需要参数解耦和整体寻优,目前常用的粒子群寻优算法普遍存在收敛速度与全局寻优能力的冲突。为此,本文基于全局最优粒子和惯性权重的关系分析,提出了全局最优比概念,并在其基础上提出了一种基于群体最优变化权重的改进粒子群算法。该算法依据迭代次数和全局最优比的大小动态调整惯性权重的值,通过动态调整惯性权重来提高算法的局部和全局寻优能力。实验表明,该算法能有效平衡粒子群算法的全局和局部搜索能力,具有较强的全局寻优能力和较快的收敛速度,不仅对单峰函数收敛效果良好,对多维多峰函数优化同样取得了较好的结果。
At present, the decoupling and overall optimization are definitely required in the optimization design process of electromechanical product system. And conflicts generally exists between the convergence speed and global optimization ability of the commonly used particle swarm optimization algorithms. Therefore, the concept of global optimal ratio is proposed in the paper, which is based on the analysis of the relationship between global optimal particle and inertia weight, and an improved particle swarm optimization algorithm is presented. According to the number of iterations and the global optimal ratio, the inertia weight is dynamically adjusted to improve the local and global optimization ability of the algorithm. Experiments show that the algorithm can effectively balance the global and local search ability of particle swarm optimization, has strong global search ability and fast convergence speed. It not only has a good convergence effect for singlepeak function, but also achieves good results for multi-multi-peak function optimization.
作者
吕振飞
李文强
徐子林
LV Zhen-fei;LI Wen-qiang;XU Zi-lin
出处
《信息技术与信息化》
2019年第12期90-94,共5页
Information Technology and Informatization
基金
科技部创新方法工作专项(2017IM040100)
四川省应用基础研究项目(No.2018JY0119)
关键词
优化设计
粒子群算法
全局最优比
惯性权重
Optimization design
particle swarm optimization
global optimal ratio
inertia weight
