期刊文献+

基于环型扩展推理规则的MaxSAT完备算法 被引量:3

MaxSAT complete algorithm based cycle extended inference rules
下载PDF
导出
摘要 最大可满足性问题(MaxSAT)是可满足性问题的优化求解问题,是经典的NP难问题.基于分支限界的MaxSAT完备算法采用推理规则、失败文字检测等方法缩短算法计算时间.推理规则产生的新子句可以构成更多的冲突集,从而有效地提高了二叉树的剪枝率和算法性能.在已有的工作基础上,针对环型结构冲突集进行分析,找到与步长大于2的环型结构冲突集等价的新子句集,并利用整数规划证明了新子句集和冲突集的MaxSAT等价性.该环型扩展推理规则产生的新3元子句亦可以提高冲突集数,提高下界.在Maxsatz2013算法的基础上实现了新算法Maxsatce.测试了MaxSAT竞赛4个类别算例集.实验结果表明环型扩展推理规则对子句长度大于等于3的MaxSAT问题,可以提高二叉树分支点的下界值,最终有效地缩减算例运算时间. Maximum Satisfiability problem(MaxSAT)is a classic optimization problem and NP-hard problem.Many NP problems can be transformed into MaxSAT problems,and many industrial problems are solved by MaxSAT algorithms effectively.Improved technologies and data structure proposed in recent years led to solve much larger instance than before.MaxSAT exact algorithm based on branch and bounds used inference rules and failed literal detection to reduce the running time.Inference rules proposed in the past time obeyed the equivalence of MaxSAT problem.Compared to not using inference rules,new clauses in clause set created by inference rules can deduce more conflict sets and improve binary tree pruning rate and algorithm's performance.Inference rules fund equivalent clauses for chain structure or cycle structure of conflict clause set.Through the analysis of cycle conflict set whose step length is greater than two,we find the new equivalent clause sets with them.The equivalent proof about new clause sets and conflict sets was described in this article.Therefore,new cycle extend inference rules were proposed to improve lower bounds and reduce the running time in further.Integer programming proved the soundness of new cycle extend inference rule.New algorithm Maxsatce adopted new rules and worked on the base of Maxsatz2013.Experiments tested four types MaxSAT instances from the website of MaxSAT competition.The experimental resultsshowed cycle extend inference rules could improve the lower bounds at binary tree branch when solving instance's clause length was greater than or equal to three.Experiment result showed cycle extend inference rule can work well when search level was close to middle of search tree.Running time of random instances and crafted instances can reduce at least 3%~14%.
出处 《南京大学学报(自然科学版)》 CAS CSCD 北大核心 2015年第4期762-771,共10页 Journal of Nanjing University(Natural Science)
基金 国家自然科学基金(51306133) 武汉科技大学校基金(2015XZ031)
关键词 NP难问题 可满足性问题 最大可满足性问题 分支限界 推理规则 环型结构 NP hard problem satisfiability problem maximum satisfiability problem branch and bounds inference rules cycle structure
  • 引文网络
  • 相关文献

参考文献17

  • 1Stephen A C. The complexity of theorem proving procedures. In:The 3rd Annual ACM Symposium on Theory of Computing. Shaker Heights, Ohio, United States : Association for Computing Machinery, 1971 : 151 - 158.
  • 2Li C, Felip M, Jordi P. New inference rules for Max-SAT. Journal of Artificial Intelligence Research, 2007,30 : 321 - 329.
  • 3Fu Z, Malik S. On solving the partial MaxSAT problem. In: The 9th Theory and Applications of Satisfiability Testing Seattle. USA: Springer Verlag,2006:252-265.
  • 4Ruben M, Vasco M, In6s L. Open-WBO: A modular MaxSAT solver. In:The 17'h Theory and Applications of Satisfiability Testing. Vienna, Austria:Springer Verlag,2014,8561:438 445.
  • 5Ruben M, Saurabh J. Incremental eardinality constraints for MaxSAT. In: The 20th International Conference on Principles and Practice of Constraint Programming. Lyon. France: Springer in the Leeture Notes in Computer Sciences Series,2014:531 48.
  • 6Argelich J, Li C, Manya F, et al. MaxSAT evaluation 2014. In: The 17th Theory and Applications of Satisfiability Testing. Vienna, Austria : Springer Verlag, 2014 : 360 - 380.
  • 7Federico H, Antonio M, Sliva J M. MaxSAT- based encodings for group MaxSAT. AI Commu- nications, 2015,28(2) : 195 - 214.
  • 8Ignatiev A, Morgado A, Manquinho V, et al. Progression in maximum satisfiability. Frontiers in Artificial Intelligenee and Applieations, 2014, 263:453 448.
  • 9Ansotegui C, Malitsky Y, Sellman M. MaxSAT by improved instance-specific algorithm configuration. In= The 28'" AAAI conference on Artificial Intelligence. Quebec city, Canada: Elserviger, 2014 : 2594- 2600.
  • 10Jessica D,Fahiem B. Exploiting the power of MIP solver in MaxSAT. In: The Applications of Saris Finland : Springer Ver I.in H, Su K. Exp fiability 16'h theory and Testing. Helsinki, lag,2013,7862 : 166-181.

二级参考文献34

  • 1李云,刘宗田,陈崚,徐晓华,程伟.多概念格的横向合并算法[J].电子学报,2004,32(11):1849-1854. 被引量:50
  • 2张文修,魏玲,祁建军.概念格的属性约简理论与方法[J].中国科学(E辑),2005,35(6):628-639. 被引量:195
  • 3凌应标,吴向军,姜云飞.基于子句权重学习的求解SAT问题的遗传算法[J].计算机学报,2005,28(9):1476-1482. 被引量:15
  • 4王道林.基于布尔矩阵的初等行变换的知识约简算法[J].计算机应用,2007,27(9):2267-2269. 被引量:9
  • 5Federico Heras, Javier Larrosa, Albert Oliveras. MINIMAX- SAT: An efficient weighted MaxSAT solver [J]. Journal of Artificial Intelligence Research, 2008, 31: 1-32.
  • 6Josep Argelieh, Li Chumin, Felip Manya, et al. Analyzing the instances of the MaxSAT evaluation [G]. LNCS 6695: Proceedings of the 14th Theory and Application of Satisfiablity Testing, 2011: 360-361.
  • 7Li Chumin, Manya Felip, Nouredine Mohamedou, et al. Re- solution based lower bounds in MaxSAT [J]. Journal of Con- straints, 2010, 15 (4): 456-484.
  • 8Li Chumin, Manya Felip, Planes Jordi. New inference rules for Max-SAT [J]. Journal of Artificial Intelligenee Research, 2007, 30: 321-329.
  • 9MaxSAT Evaluation Website [CP/OL]. http: //www. max- sat. udl. cat, 2013.
  • 10Carlos Ansotegui, Maria Luisa Eonet, Jordi Levy. A new algo- rithms for weighted partial MaxSAT [C] //Proceedings of the 24th AAAI Conference on Artificial Intelligence, 2010: 3-8.

共引文献19

同被引文献11

引证文献3

二级引证文献9

相关主题

;
使用帮助 返回顶部