华南理工大学学报(自然科学版) ›› 2009, Vol. 37 ›› Issue (9): 112-116.

• 机械工程 • 上一篇    下一篇

一种用于车间调度的基于熵的混合遗传算法

陈耀军 姚锡凡 张庆   

  1. 华南理工大学 机械与汽车工程学院, 广东 广州 510640
  • 收稿日期:2008-12-30 修回日期:2009-02-19 出版日期:2009-09-25 发布日期:2009-09-25
  • 通信作者: 陈耀军(1979-),男,博士,主要从事制造系统控制与调度优化研究. E-mail:butianshi00@163.com
  • 作者简介:陈耀军(1979-),男,博士,主要从事制造系统控制与调度优化研究.
  • 基金资助:

    国家“863”高技术计划项目(2007AA04Z111)

An Entropy-Based Hybrid Genetic Algorithm for Job-Shop Scheduling

Chen Yao-jun  Yao Xi-fan  Zhang Qing   

  1. School of Mechanical and Automotive Engineering, South China University of Technology, Guangzhou 510640, Guangdong, China
  • Received:2008-12-30 Revised:2009-02-19 Online:2009-09-25 Published:2009-09-25
  • Contact: 陈耀军(1979-),男,博士,主要从事制造系统控制与调度优化研究. E-mail:butianshi00@163.com
  • About author:陈耀军(1979-),男,博士,主要从事制造系统控制与调度优化研究.
  • Supported by:

    国家“863”高技术计划项目(2007AA04Z111)

摘要: 为提高车间调度算法的寻优性能,通过对模拟退火遗传算法收敛图的研究,提出了评价算法种群有序性(差异性)的种群熵.基于种群熵,提出了改进的模拟退火遗传算法.该混合算法通过种群熵动态地改变算法的交叉和变异概率,使之适应种群的变化,提高种群个体的多样性,克服算法的过早收敛,从而达到提高算法寻优性能的目的.仿真实例表明:所提出的算法的寻优性能有了显著的提高;与传统遗传算法相比,在相同的初始条件与收敛条件下,对于FT10问题,最优结果由998优化至930,偏差由9.78%减少到了1.94%;对于LA36问题,最优结果由1359优化至1278,偏差由10.29%减少到了2.97%.

关键词: 遗传算法, 模拟退火算法, 种群熵, 车间调度, 混合算法

Abstract:

To improve the searching performance of job-shop scheduling algorithms,the population entropy for eva-luating the orderliness(difference) of the algorithm population is proposed by analyzing the convergent graph of the simulated annealing genetic algorithm.Then,a modified simulated annealing genetic algorithm is proposed based on the population entropy,which adapts itself to the variation of population via the dynamic change of both the crossover probability and the mutation probability with the population entropy. Thus, the population diversity is increased, the premature convergence of the algorithm is overcome, and the searching performance is improved. Simulated results show that the proposed algorithm has great superiority in the searching performance. As compared with the traditional genetic algorithm, in the same initial and convergence conditions for the FT10 problem, the optimal result is improved from 998 to 930 and the deviation reduces from 9.78% to 1.94%, while for the LA36 problem, the optimal result is improved from 1 359 to 1 278 and the deviation reduces from 10. 29% to 2.97%.

Key words: genetic algorithm, simulated annealing algorithm, population entropy, job-shop scheduling, hybrid algorithm