J4 ›› 2013, Vol. 51 ›› Issue (02): 257-266.

• 计算机科学 • 上一篇    下一篇

基于分割的超树分解方法

王涛1, 王瑞芹2,3, 李占山2,3, |陈超2,3   

  1. 1. 长春工业大学 计算机科学与工程学院, 长春 130012;
    2. 吉林大学 符号计算与知识工程教育部重点实验室, 长春 130012;
    3. 吉林大学 计算机科学与技术学院, 长春 130012
  • 收稿日期:2012-05-02 出版日期:2013-03-26 发布日期:2013-03-27
  • 通讯作者: 李占山 E-mail:zslizsli@163.com

SeparationBased Hypertree Decomposition

WANG Tao1, WANG Ruiqin2,3, LI Zhanshan2,3, CHEN Chao2,3   

  1. 1. School of Computer Science and Engineering, Changchun University of Technology, Changchun 130012, China;
    2. Key Laboratory of Symbolic Computation and Knowledge Engineering for Ministry of Education, Jilin University, Changchun 130012, China;
    3. College of Computer Science and Technology, Jilin University, Changchun 130012, China
  • Received:2012-05-02 Online:2013-03-26 Published:2013-03-27
  • Contact: LI Zhanshan E-mail:zslizsli@163.com

摘要:

基于det-k-decomp算法, 通过引入同构的概念和对separator选择空间的进一步限制, 提出一类新的超树分解: 分割的超树分解, 并提出一种具有较小超树宽度的超树分解方法: 基于分割的超树分解——sht-k-decomp, 该算法能有效提高约束满足问题的求解效率. 实验结果表明, sht-k-decomp算法多数情况下效率高于det-k-decomp算法.

关键词: 人工智能; 超树分解; 约束满足问题

Abstract:

Based on det-k-decomp algorithm, along with the noting  isomorphic component and reducing the search space for an appropriate separator, the authors  presented a new class of hypertree decomposition: separated hypertree decomposition and presented a new tractable hypertree decomposition algorithm——sht-k-decomp, which can improve the efficiency of constrain satisfaction problem solving effectively. The experimental results on most instances show that our method outperforms det-k-decomp.

Key words: artificial intelligence, hypertree decomposition, constraint satisfaction problems

中图分类号: 

  • TP18