Heuristics for planning with SAT

  • Authors:
  • Jussi Rintanen

  • Affiliations:
  • NICTA and the Australian National University, Canberra, Australia

  • Venue:
  • CP'10 Proceedings of the 16th international conference on Principles and practice of constraint programming
  • Year:
  • 2010

Quantified Score

Hi-index 0.00

Visualization

Abstract

Generic SAT solvers have been very successful in solving hard combinatorial problems in various application areas, including AI planning. There is potential for improved performance by making the SAT solving process more application-specific. In this paper we propose a variable selection strategy for AI planning. The strategy is based on generic principles about properties of plans, and its performance with standard planning benchmarks often substantially improves on generic variable selection heuristics used in SAT solving, such as the VSIDS strategy. These improvements lift the efficiency of SAT based planning to the same level as best planners that use other search methods.