• 论文 •    

求解装配序列规划的一种多智能体进化算法

曾聪文,古天龙   

  1. 1.西安电子科技大学 电子工程学院,陕西西安710071;2.桂林电子科技大学 计算机与控制学院,广西桂林541004
  • 出版日期:2009-09-15 发布日期:2009-09-25

Multi-Agent evolutionary algorithm for assembly sequence planning

ZENG Cong-wen, GU Tian-long   

  1. 1.School of Electronic Engineering, Xidian University, Xian 710071, China;2.School of Computer Science, Guilin University of Electronic Technology, Guilin 541004, ChinaTo improve the individuals intelligence of traditional evolutionary algorithms for Assembly Sequence Planning (ASP, a multi-Agent evolutionary algorithm for ASP was presented. Agent which carried a possible assembly sequence was not only an basic evolutionary operation units,but also with such intelligent actions as learning
  • Online:2009-09-15 Published:2009-09-25

摘要: 针对求解装配序列规划的进化类算法中个体智能性不足的问题,设计了求解装配序列规划的一种多智能体进化算法。携有装配序列规划信息的智能体不仅是竞争、变异等进化操作的基本单元,同时又具有“学习”、“协作”等智能行为。实验表明,与其他各类进化算法相比,多智能体进化算法具有明显的优越性。

关键词: 智能体, 学习, 协作, 进化算法, 装配序列规划

Abstract: To improve the individuals intelligence of traditional evolutionary algorithms for Assembly Sequence Planning (ASP), a multi-Agent evolutionary algorithm for ASP was presented. Agent which carried a possible assembly sequence was not only an basic evolutionary operation units,but also with such intelligent actions as learning and cooperation. Experimental results showed that the algorithm was superior to other evolutionary algorithms.

Key words: Agent, learning, coorperation, evolutionary algorithm, assembly sequences planning

中图分类号: