摘要
针对含多挤压油膜阻尼器-转子系统的双稳态解求解问题,根据挤压油膜阻尼器动力特性系数特点,提出了一种混合式聚类粒子群算法用于搜索求解系统的解。在本算法中,每个粒子不再关注全局最优粒子,而是关注由粒子的位置满意度和视距所确定的学习集群的最优粒子。同时,将所有由相互关注的粒子定义为一个目标集群,从而实现全体粒子的聚类分群。每个满足解条件的目标集群的最优粒子即为一个潜在的解。为了加快迭代速度,对每个目标集群的最优粒子按照牛顿迭代法进行搜索。案例分析表明该聚类算法可以有效实现粒子的分群,求解出挤压油膜阻尼器-转子系统的双稳态解。
In this paper,a hybrid clustering Particle Swarm Optimization(PSO)algorithm is proposed based on the characteristics of the dynamic characteristic coefficients of the extrusion oil film damper,to search and solve the bistable solution of a rotor system with multiple oil film dampers.For this algorithm,each particle is no longer concerned with the global optimal particle,but with the optimal particle of the learning cluster determined by the particle′s position satisfaction and visual distance.At the same time,all particles concerned by each other are defined as a target cluster,so as to achieve the clustering of particles.The optimal particle of each target cluster that satisfies the solution condition is a solution.In order to speed up the iteration,the optimal particles of each target cluster are searched according to Newton iteration method.From the results of case study and discussion,it can be seen that the clustering algorithm proposed in this paper can effectively realize particle clustering and effectively solve the bistable solution of the oil film damp-rotor system.
作者
邱玉江
李书鹏
刘品潇
艾三
徐沛
QIU Yu-jiang;LI Shu-peng;LIU Pin-xiao;AI San;XU Pei(Academy of Mechanical Discipline Studies,Nanyang Institute of Technology,Nanyang 473004,China;Jinguan Electric Co Ltd,Nanyang 473003,China)
出处
《南阳理工学院学报》
2025年第4期48-53,78,共7页
Journal of Nanyang Institute of Technology
基金
南阳市科技计划项目(24KJGG083)
河南省高等学校重点科研项目(25A460024)
南阳理工学院大学生科研基金项目(NGDJ-2024-16)。
关键词
挤压油膜阻尼器
双稳态解
粒子群算法
聚类算法
extruded oil film damper
bistable solution
particle swarm optimization
clustering algorithm
作者简介
邱玉江(1978-),男,博士,讲师,主要研究方向:机械系统动力学,数字化设计。
