期刊文献+

利用派生谓词和偏好处理OSP问题的目标效益依赖

Handling Goal Utility Dependencies in OSP Problems with Derived Predicates and Preferences
下载PDF
导出
摘要 在过度规划问题(over-subscribed planning,简称OSP)研究中,如果目标之间不是相互独立的,那么目标间的效益依赖比单个目标效益更能提高规划解的质量.但是已有的描述模型不符合标准规划描述语言(planningdomain description language,简称PDDL)的语法规范,不能在一般的OSP规划系统上进行推广.提出了用派生谓词规则和目标偏好描述效益依赖的方法,这二者均为PDDL语言的基本要素.实质上,将已有的GAI模型转换为派生谓词规则和目标偏好,其中派生谓词规则显式描述目标子集的存在条件,偏好机制用来表示目标子集的效益,二者缺一不可.该转换算法既可以保持在描述依赖关系时GAI模型的易用性和直观性上,又可以扩展一般的OSP规划系统处理目标效益依赖的能力.从理论上可以证明该算法在转换过程中的语义不变性,在基准领域的实验结果表明其可行性和对规划解质量的改善能力.提出符合PDDL语言规范的目标效益依赖关系的描述形式,克服了已有模型不通用的缺点. In the field of over-subscribed planning (OSP), goal utility dependencies are more useful than a single goal utility used to improve the plan quality, if goals are not independent. However, existing description models do not follow the grammatical specification of standard planning domain description language (PDDL), so they cannot be used in other OSP planning systems yet. To solve this, this paper presents a new way of describing goal utilitydependencies with derived predicate rules and goal preferences, both of which are essential elements of PDDL. The goal of the process is to transform GAI (general additive independence) models into these two elements, where a derived predicate rule is used to describe the explicitly triggering conditions of a goal sub-set. A preference is usedto depict explicitly its utility or value and both are indispensable. This compilation mechanism can not only maintain the characteristic of ease-of-use and straightness of GAI models in describing utility dependencies, but can also expend the ability of handling utility dependencies for general OSP planning systems. Also, this paper provesthe semantic conservation in the compilation process. Experimental results in some OSP benchmark domains show that the algorithm is feasible and useful for improving the plan quality. It is the first time to describe goal utility dependency with PDDL elements in order to overcome the limitations of existing models.
出处 《软件学报》 EI CSCD 北大核心 2012年第3期439-450,共12页 Journal of Software
基金 国家自然科学基金(61100134 61003179 60903178) 广东省自然科学基金(S2011040001427)
关键词 人工智能 智能规划 过度规划问题 目标效益依赖 派生谓词规则 偏好 artificial intelligence (AI) automated planning over-subscribed planning (OSP) goal utilitydependency derived predicate rules preference
  • 相关文献

参考文献26

  • 1Edelkamp S,Jabbar S,Nazih M.Large-Scale optimal PDDL3 planning with MIPS-XXL.In:Long D,Smith SF,Borrajo D,McCluskey L,eds.Proc.of the 16th Int’l Conf.on Automated Planning and Scheduling.Menlo Park:AAAI,2006.81-92.
  • 2Edelkamp S,Jabbar S.Cost-Optimal external planning.In:Proc.of the 21st National Conf.on Artificial Intelligence.Menlo Park:AAAI,2006.821-826.
  • 3Ghallab M,Nau D,Traverso P.Automated Planning:Theory and Practice.San Fransisco:Morgan Kaufmann Publishers,2004.1-663.
  • 4Coles AI,Smith AJ.Marvin:Macroactions from reduced versions of the instance.In:Booklet of the 4th Int’l Planning Competition.2004.24-26.http://www.tzi.de/~edelkamp/ipc-4/publication.html.
  • 5Gerevini A,Saetti A,Serina I,Toninelli P.Fast planning in domains with derived predicates:An approach based on rule-action graphs and local search.In:Veloso MM,Kambhampati S,eds.Proc.of the 20th National Conf.on Artificial Intelligence.Menlo Park:AAAI,2005.1157-1162.
  • 6Sohrabi S,Baier JA,McIlraith SA.HTN planning with preferences.In:Boutilier C,ed.Proc.of the 21st Int’l Joint Conf.on Artificial Intelligence.Menlo Park:AAAI,2009.1790-1797.
  • 7Hsu CW,Wah BW.The SGPlan planning system in IPC-6.In:Booklet of the Deterministic Part of the 6th Int’l Planning Competition.2008.1-3.http://ipc.informatik.uni-freiburg.de/Planners.
  • 8Edelkamp S,Kissmann P.Optimal symbolic planning with action costs and preferences.In:Boutilier C,ed.Proc.of the 21st Int’l Joint Conf.on Artificial Intelligence.Menlo Park:AAAI,2009.1690-1695.
  • 9Nigenda RS,Kambhampati S.Planning graph heuristics for selecting objectives in over-subscription planning problems.In:Biundo S,Myers KL,Rajan K,eds.Proc.of the 15th Int’l Conf.on Automated Planning and Scheduling.Menlo Park:AAAI,2005.192-201.
  • 10Li L,Onder N.Generating plans in concurrent,probabilistic,over-subscribed domains.In:Fox D,Gomes CP,eds.Proc.of the 23rd National Conf.on Artificial Intelligence.Menlo Park:AAAI,2008.1857-1858.

二级参考文献50

  • 1蒋志华,姜云飞.基于与状态无关的激活集的包含派生谓词的规划问题求解[J].计算机科学,2007,34(3):176-180. 被引量:3
  • 2Thiebaux S, Hoffmann J, Nebel B. In defense of PDDL axioms. Artifieial Intelligence, 2005, 168(1/2) : 38-69.
  • 3Fikes R, Nilsson N. STRIPS: A new approach to the application of theorem proving to problem solving. Artificial Intelligence, 1971, 2(3/4) : 189-208.
  • 4Thielscher M. Ramification and causality. Artificial Intelligence, 1997, 89(1/2):317-364.
  • 5Yang Qing, Wu Kang-Hen, Jiang Yun-Fei. Learning action models from plan examples using weighted MAX-SAT. Artificial Intelligence, 2007, 171(2-3): 107-143.
  • 6Amir E. Learning partially observable deterministic action models//Proceedings of the International Joint Conference on Artificial Intelligence (IJCAI 2005). Edinburgh, Scotland, UK, 2005:1433-1439.
  • 7Ilghami O, Munoz-Avila H, Nau D S, Aha D W. Learning preconditions for planning from plan traces and HTN struc ture. Journal of Artificial Intelligence Research, 2005, 21 (4) : 388-413.
  • 8Aler R, Borrajo D, Isasi P. Using genetic programming to learn and improve control knowledge. Artificial Intelligence, 2002, 141(1/2): 29-56.
  • 9Huang Y C, Selman B, Kautz H A. Learning declarative control rules for constraint-based planning//Proceedings of International Conference on Machine Learning. Stanford, CA, USA, Morgan Kaufmann, 2000:415-422.
  • 10Bianehi R A C, Ribeiro C H C, Costa A H R. Heuristic selection of actions in multiagent reinforcement learning//Proceedings of the IJCAI 2007. Hyderabad, 2007:690-695.

共引文献18

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部