Plan-refinement strategies and search-space size
During the planning process, a planner may have many options for refinements to perform on the plan being developed. The planner's efficiency depends on how it chooses which refinement to do next. Recent studies have shown that several versions of the popular “least commitment” plan refinement strategy are often outperformed by a fewest alternatives first (FAF) strategy that chooses to refine the plan element that has the smallest number of alternative refinement options.
A refinement planner's search space is an AND/OR graph, and the planner “serializes” this graph by mapping it into an equivalent state-space graph. Different plan refinement strategies produce different serializations of the AND/OR graph.
The sizes of different serializations of the AND/OR graph can differ by an exponential amount. A planner whose refinement strategy produces a small serialization is likely to be more efficient than a planner whose refinement strategy produces a large serialization.
The FAF heuristic can be computed in constant time, and in our experimental studies it usually produced an optimal or near-optimal serialization. This suggests that using FAF (or some similar heuristic) is preferable to trying to guarantee an optimal serialization (which we conjecture is a computationally intractible problem).
Keywordsplanning and search refinement strategies commitment strategies
Unable to display preview. Download preview PDF.
- [Barret, 1994][Barret and Weld, 1994] Anthony Barret and Daniel Weld. Partial-order planning: Evaluating possible efficiency gains. Artificial Intelligence 67(1), pp. 71–112Google Scholar
- [Bitner, 1975][Bitner and Reingold, 1975] James Bitner and Edward Reingold. Backtrack Programming Techniques. CACM 18(11), pp. 651–656.Google Scholar
- [Clocksin and Mellish, 1981]W. Clocksin and C. Mellish. Programming in PROLOG. Springer-Verlag.Google Scholar
- [Currie, 1991][Currie and Tate, 1991] Ken Currie and Austin Tate. O-plan: the open planning architecture. Artificial Intelligence 52, pp. 49–86.Google Scholar
- [Erol, 1995]Kutluhan Erol. HTN planning: Formalization, analysis, and implementation. Ph.D. dissertation, Computer Science Dept., U. of Maryland.Google Scholar
- [Gupta, 1992][Gupta and Nau, 1992] Naresh Gupta and Dana Nau. On the complexity of blocksworld planning. Artificial Intelligence 56:2-3, pp. 223–254.Google Scholar
- [Gerevini, 1996][Gerevini and Schubert, 1996] Alfonso Gerevini and Lenhart Schubert. Accelerating Partial-Order Planners: Some Techniques for Effective Search Control and Pruning. Journal of Artificial Intelligence Research 5, pp. 95–137.Google Scholar
- [Joslin and Pollack, 1994]David Joslin and Martha Pollack. Least-cost flaw repair: A plan refinement strategy for partial-order planning. In Proceedings of the Twelfth National Conference on Artificial Intelligence, pp. 1004–1009.Google Scholar
- [Joslin and Pollack, 1996]David Joslin and Martha Pollack. Is “early commitment” in plan generation ever a good idea? In Proc. Thirteenth National Conference on Artificial Intelligence, pp. 1188–1193.Google Scholar
- [Kabhampati, 1995][Kambhampati et al., 1995] Subbarao Kambhampati, Craig Knoblock, and Qiang Yang. Planning as refinement search: A unified framework for evaluating design tradeoffs in partial-order planning. Artificial Intelligence 76, pp. 167–238.Google Scholar
- [Kumar, 1992]Vipin Kumar. Algorithms for constraint-satisfaction problems: A survey. AI Magazine, pp. 32–44.Google Scholar
- [Luke, 1997]Sean Luke. A Fast Probabilistic Tree Generation Algorithm. Unpublished manuscript.Google Scholar
- [Penberthy and Weld, 1992]J. S. Penberthy and Daniel Weld. UCPOP: A sound, complete, partial order planner for ADL. Proc. KR-92.Google Scholar
- [Purdom, 1983][Purdom, 1983] Paul W. Purdom. Search Rearrangement Backtracking and Polynomial Average Time. Artificial Intelligence 21, pp. 117–133.Google Scholar
- [Purdom, 1983][Purdom and Brown, 1983] Paul W. Purdom and Cynthia A. Brown. An Analysis of Backtracking with Search Rearragement. SIAM J. Computing 12(4), pp.717–733Google Scholar
- [Sacerdoti, 1977]Earl Sacerdoti. A Structure for Plans and Behavior. American Elsevier Publishing Company.Google Scholar
- [Stefik, 1981][Stefik, 1981] Mark Stefik. Planning with constraints (MOLGEN: part 1). Artificial Intelligence 16, pp. 111–140.Google Scholar
- [Smith et al., 1996]S. J. J. Smith, D. S. Nau, and T. A. Throop. Total-order multiagent task-network planning for control bridge. AAAI-96, pp. 108-113.Google Scholar
- [Stockman, 1979][Stockman, 1979] G. Stockman. A minimax algorithm better than alpha-beta? Artificial Intelligence 12(2), pp. 179–96.Google Scholar
- [Tsuneto, 1996][Tsuneto et al., 1996] Reiko Tsuneto, Kutluhan Erol, James Hendler, and Dana Nau. Commitment strategies in hierarchical task network planning. In Proc. Thirteenth National Conference on Artificial Intelligence, pp. 536–542.Google Scholar
- [Veloso, 1995][Veloso and Stone, 1995] Manuela Veloso and Peter Stone. FLECS: Planning with a flexible commitment strategy. JAIR 3, pp. 25–52.Google Scholar