计算物理 ›› 1994, Vol. 11 ›› Issue (3): 278-282.

• 论文 • 上一篇    下一篇

推销员问题的重要抽样模拟退火方法

陈军, 陈天崙, 黄五群   

  1. 南开大学物理系, 天津 300071
  • 收稿日期:1992-10-30 修回日期:1993-10-16 出版日期:1994-09-25 发布日期:1994-09-25
  • 基金资助:
    国家基础性研究重大项目《非线性科学》及国家自然科学基金

THE TRAVELING SALESMAN PROBLEM: OPTIMIZATION BY IMPORTANCE SAMPLING SIMULATED ANNEALING METHOD

Chen Jun, Chen Tianlun, Huang Wuqun   

  1. Dapartment of Physics, Nankai University Tianjin 300071
  • Received:1992-10-30 Revised:1993-10-16 Online:1994-09-25 Published:1994-09-25

摘要: 采用随机三角点阵上城市间的最近邻关系,构造路径子空间来求解旅行推销员问题。用重要抽样的模拟退火算法及分段优化法大大提高了计算的效率,节省了计算时间,得到较优的结果。

关键词: 随机三角点阵, 重要抽样, 模拟退火, 旅行推销员问题

Abstract: The nearest neighbour relation between cities on random triangle lattice has been used to construct a tour subspace and to solve the traveling salesman problem. With importance sampling simulated annealing method and subtour optimization method, the efficiency has been raised obviously. Near optimal solutions are obtained in shorter computation time.

Key words: random triangle lattice, importance sampling, simulated annealing, traveling salesman problem

中图分类号: