• 论文 •    

具有资源约束的项目调度问题中的量子进化算法

黄志宇   

  1. 中国科学院 自动化研究所复杂系统与智能科学实验室,北京100080
  • 出版日期:2009-09-15 发布日期:2009-09-25

Quantum-inspired evolutionary algorithm for resources constrainted project scheduling problem

HUANG Zhi-yu   

  1. Lab of Complex Systems and Intelligence Science, Institute of automation, Chinese Academy of Sciences, Beijing 100080, China
  • Online:2009-09-15 Published:2009-09-25

摘要: 为了将量子进化思想应用到具有资源约束的项目调度领域中,提出了一种基于活动对连接关系的调度解表示方法,进而构成新的搜索解空间。利用新表示方法的二值性,给出了量子进化算法。该算法扩展了前向-后向搜索思想,构造出原问题的反向问题,并把原问题和反向问题以同样的方式进行求解,利用反向搜索一个方向上的解群来更新另一个方向上的解群。此外,算法根据分散搜索的思想,利用新搜索解空间中定义的距离来控制搜索的分散性。最后,通过仿真说明了该算法的有效性。

关键词: 项目调度, 分散搜索, 量子进化算法, 反向调度

Abstract: To apply quantum-inspired evolutionary into the field of the Resources Constrained Project Scheduling Problem (RCPSP), a new representation of the RCPSP solution was proposed which was based on the connections of every activities pair. And the new search space was generated. Based on the two-value property of the new representation, a quantum-inspired evolutionary algorithm was put forward. In this new algorithm, the idea of forward-backward scheduling was extended and a reverse problem was generated. The original and reverse problems were solved by the same means. For linking the two direction problems, a reverse scheduling process was used, in which, one direction population was reverse scheduled to update the other direction population. Furthermore, based on the idea of scatter search, the definition of distance between two solutions was used to control the dispersion of the search space. The effectiveness of the new algorithm was demonstrated by the simulation.

Key words: project scheduling, scatter search, quantum-inspired evolutionary algorithm, reverse scheduling

中图分类号: