首页> 外文会议>International conference on artificial intelligence;ICAI 2011 >Yet another breakout inspired infeasible subset detection in constraint satisfaction problem
【24h】

Yet another breakout inspired infeasible subset detection in constraint satisfaction problem

机译:另一个突破激发了约束满足问题中不可行的子集检测

获取原文
获取外文期刊封面目录资料

摘要

In [1], Eisenberg and Faltings proposed a breakout [2] based hybrid approach to detect infeasible subset (IS) in constraint satisfaction problems. In this paper, we adopted this algorithm as a preprocessing procedure to reduce the over-constrained constraint satisfaction problem. Then a three procedures hybrid approach is employed to detect the irreducible infeasible subset (IIS) of identified IS. The approach consists of three procedures - critical constraint locating, critical constraint extending and critical constraint subsets satisfiability testing. The extent of the subproblem is based on the first located critical constraint and is ended while it cannot be satisfied by a Tabu Search incomplete search algorithm. Finally, the MAC (Maintaining Arc Consistency during search) method provides the infeasibility of the extracted subproblem. By adopting a hybrid approach, the tractability of satisfiability testing of real world applications becomes realistic. The experimental results demonstrate the effectiveness of the method.
机译:在[1]中,Eisenberg和Faltings提出了一种基于突破[2]的混合方法来检测约束满足问题中的不可行子集(IS)。在本文中,我们采用该算法作为预处理程序,以减少过度约束的约束满足问题。然后,采用三种程序的混合方法来检测已识别IS的不可约的不可行子集(IIS)。该方法包括三个过程-关键约束定位,关键约束扩展和关键约束子集可满足性测试。子问题的程度基于第一个定位的关键约束,并且在禁忌搜索不完全搜索算法无法满足的情况下结束。最后,MAC(在搜索过程中保持弧线一致性)方法提供了所提取子问题的不可行性。通过采用混合方法,现实应用程序的可满足性测试的可处理性变得现实。实验结果证明了该方法的有效性。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号