• 论文 •    

基于改进蚁群算法的装配序列规划

史士财,李荣,付宜利,马玉林   

  1. 1.哈尔滨工业大学 机器人技术与系统国家重点实验室,黑龙江哈尔滨150001;2.黑龙江工程学院 汽车系,黑龙江哈尔滨150050;3.哈尔滨工业大学 现代生产技术中心,黑龙江哈尔滨150001
  • 出版日期:2010-06-15 发布日期:2010-06-25

Assembly sequence planning based on improved ant colony algorithm

SHI Shi-cai, LI Rong, FU Yi-li, MA Yu-lin   

  1. 1.State Key Laboratory of Robots & System, Harbin Institute of Technology, Harbin 150001, China; 2.Department of Automobile, Heilongjiang Institute of Technology, Harbin 150050, China; 3.Advanced Manufacturing Technology Center, Harbin Institute of Technology, Harbin 150001, China
  • Online:2010-06-15 Published:2010-06-25

摘要: 针对装配序列规划问题,分析了基本蚁群系统的不足,提出了面向装配序列规划的改进蚁群算法,来获得最优或次最优的装配序列。改进蚁群算法中,将装配操作约束作为启发式信息引入状态转移概率中,通过获取零部件之间的装配关系设定可行转移范围。通过信息素残留系数的动态变化和影响转移概率的α、β参数的动态设置,提高了蚁群的收敛速度并有效地避免了其陷入局部最优解。通过实例验证了改进算法的有效性。

关键词: 蚁群算法, 装配序列规划, 信息素

Abstract: Aiming at Assembly Sequence Planning(ASP)problem, shortcomings of basic ant system were analyzed, an improved Ant Colony Algorithm(ACA)oriented to ASP was proposed to obtain optimal or near optimal assembly sequence. In this algorithm, assembly operation constraint was introduced into the state transfer function as heuristic information. And feasible transition area was set up by obtaining assembly relationship of the parts. By dynamic change of pheromohe trail persistence and dynamic setting of parameters α and β, the convergence speed of ACA was improved and the local optimization was avoided. Finally, the effectiveness was verified by an example.

Key words: ant colony algorithm, assembly sequence planning, pheromone, optimization

中图分类号: