Phase Transitions in Classical Planning: An Experimental Study

Phase transitions in the solubility of problem instances are known in many types of computational problems relevant for artificial intelligence, most notably for the satisfiability problem of the classical propositional logic. However, phase transitions in classical planning have received far less attention. Bylander has investigated phase transitions theoretically as well as experimentally by using simplified planning algorithms, and shown that most of the soluble problems can be solved by a naive hill-climbing algorithm. Because of the simplicity of his algorithms he did not investigate hard problems on the phase transition region. In this paper, we address exactly this problem. We introduce two new models of problem instances, one eliminating the most trivially insoluble instances from Bylander's model, and the other restricting the class of problem instances further. Then we perform experiments on the behavior of different types of planning algorithms on hard problems from the phase transition region, showing that a planner based on general-purpose satisfiability algorithms outperforms two planners based on heuristic local search.

[1]  Bart Selman,et al.  Heavy-Tailed Phenomena in Satisfiability and Constraint Satisfaction Problems , 2000, Journal of Automated Reasoning.

[2]  Jussi Rintanen,et al.  A Planning Algorithm not based on Directional Search , 1998, KR.

[3]  Pekka Orponen,et al.  An efficient local search method for random 3-satisfiability , 2003, Electron. Notes Discret. Math..

[4]  Lawrence Ryan.  Efficient algorithms for clause-learning SAT solvers , 2004 .

[5]  John K. Slaney,et al.  On the Hardness of Decision and Optimisation Problems , 1998, ECAI.

[6]  M. Fox,et al.  The 3rd International Planning Competition: Results and Analysis , 2003, J. Artif. Intell. Res..

[7]  Ilkka Niemelä,et al.  Parallel Encodings of Classical Planning as Satisfiability , 2004, JELIA.

[8]  Karem A. Sakallah,et al.  Computing Vertex Eccentricity in Exponentially Large Graphs: QBF Formulation and Solution , 2003, SAT.

[9]  Tom Bylander,et al.  The Computational Complexity of Propositional STRIPS Planning , 1994, Artif. Intell..

[10]  Avrim Blum,et al.  Fast Planning Through Planning Graph Analysis , 1995, IJCAI.

[11]  M. Mézard,et al.  Analytic and Algorithmic Solution of Random Satisfiability Problems , 2002, Science.

[12]  Bernhard Nebel,et al.  The FF Planning System: Fast Plan Generation Through Heuristic Search , 2011, J. Artif. Intell. Res..

[13]  Bart Selman,et al.  Pushing the Envelope: Planning, Propositional Logic and Stochastic Search , 1996, AAAI/IAAI, Vol. 2.

[14]  Peter C. Cheeseman,et al.  Where the Really Hard Problems Are , 1991, IJCAI.

[15]  Drew McDermott,et al.  The 1998 AI Planning Systems Competition , 2000, AI Mag..

[16]  Toby Walsh,et al.  Beyond NP: the QSAT phase transition , 1999, AAAI/IAAI.

[17]  Hector J. Levesque,et al.  Generating Hard Satisfiability Problems , 1996, Artif. Intell..

[18]  Tad Hogg,et al.  Phase Transitions in Artificial Intelligence Systems , 1987, Artif. Intell..

[19]  Blai Bonet,et al.  Planning as heuristic search , 2001, Artif. Intell..

[20]  Jussi Rintanen,et al.  Symmetry Reduction for SAT Representations of Transition Systems , 2003, ICAPS.

[21]  Béla Bollobás,et al.  Random Graphs , 1985 .

[22]  Ivan Serina,et al.  LPG: A Planner Based on Local Search for Planning Graphs with Action Costs , 2002, AIPS.

[23]  Gilles Dequen,et al.  A backbone-search heuristic for efficient solving of hard 3-SAT formulae , 2001, IJCAI.

[24]  Tom Bylander,et al.  A Probabilistic Analysis of Propositional STRIPS Planning , 1996, Artif. Intell..

[25]  Kenneth L. McMillan,et al.  Interpolation and SAT-Based Model Checking , 2003, CAV.