Chapter 7 - Propositional Satisfiability Techniques

Abstract

The general idea underlying the “planning as satisfiability” approach is to map a planning problem to a well-known problem for which there exists effective algorithms and procedures. A plan is then extracted from the solution to the new problem. More specifically, the idea is to formulate a planning problem as a propositional satisfiability problem—that is, as the problem of determining whether a propositional formula is satisfiable. Recent improvements in the performance of general purpose algorithms for propositional satisfiability provide the ability to scale up to relatively large problems. The ability to exploit the efforts and the results in a very active field of research in computer science is indeed one of the main advantages of this approach. This chapter focuses on the encoding of planning problems into satisfiability problems and describes some existing satisfiability procedures that have been used in planning. This chapter discusses a way to translate a planning problem to a propositional formula. The chapter shows how standard decision procedures can be used as planning algorithms. The chapter discusses some different ways to encode a planning problem into a satisfiability problem.