【24h】

Ordering heuristics for Singleton Arc Consistency

机译:单例弧一致性的排序试探法

获取原文

摘要

The use of consistency techniques is the main feature of any Constraint Satisfaction Problems(CSPs)solver. Arc Consistency (AC) applied to preprocessing can reduce the search space by removingthe values which are arc inconsistent. Singleton Arc Consistency (SAC) is a stronger level of consistencythan AC. Arc consistency technique with heuristic strategy has been proved to be an efficient techniqueto improve the efficiency of algorithms. In this paper, we first study the well-known SAC-3 algorithm indetail, and propose two algorithms with heuristic strategies for the drawbacks in SAC-3, SAC-3-FFPand SAC-3-MINIsup. Then, we give the correctness proof and complexity analysis. In the end, we implement the two algorithms in some random CSPs and benchmarks during the preprocessing step. Experimental results show that the two algorithms perform better than SAC-3 for most test cases.
机译:一致性技术的使用是任何约束满足问题(CSP)解决方案的主要特征。应用于预处理的弧一致性(AC)可以通过删除弧不一致的值来减少搜索空间。单例电弧一致性(SAC)比AC具有更高的一致性。实践证明,采用启发式策略的弧一致性技术是提高算法效率的有效技术。在本文中,我们首先详细研究了著名的SAC-3算法,并针对SAC-3的缺点提出了两种具有启发式策略的算法:SAC-3-FFP和SAC-3-MINIsup。然后,我们给出了正确性证明和复杂性分析。最后,在预处理步骤中,我们在一些随机CSP和基准中实现了这两种算法。实验结果表明,在大多数测试案例中,这两种算法的性能均优于SAC-3。

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