Concise finite-domain representations for PDDL planning tasks

Abstract

We introduce an efficient method for translating planning tasks specified in the standard PDDL formalism into a concise grounded representation that uses finite-domain state variables instead of the straight-forward propositional encoding.Translation is performed in four stages. Firstly, we transform the input task into an equivalent normal form expressed in a restricted fragment of PDDL. Secondly, we synthesize invariants of the planning task that identify groups of mutually exclusive propositions which can be represented by a single finite-domain variable. Thirdly, we perform an efficient relaxed reachability analysis using logic programming techniques to obtain a grounded representation of the input. Finally, we combine the results of the third and fourth stage to generate the final grounded finite-domain representation.The presented approach has originally been implemented as part of the Fast Downward planning system for the 4th International Planning Competition (IPC4). Since then, it has been used in a number of other contexts with considerable success, and the use of concise finite-domain representations has become a common feature of state-of-the-art planners.

Using random sampling trees for automated planning

“…Another method for finding binary mutexes is the monotonicity analysis generally employed to generate a multi-valued formalization of the problem [22]. This monotonicity analysis ensures that the number of propositions true at the same time that belong to a set I g = {p 0 , p 1 , .…”

A hybrid bit-encoding for SAT planning based on clique-partitioning

“…Planning has PSPACE complexity [3].In the paper, we use the Planning Domain Definition Language (PDDL) to define planning problems. To the best of our knowledge, PDDL the most extended language for automated planning [4,5,6,7].…”

“…Actions are also grounded and encoded as logic variables in a similar way. The process can generate an explosion of logical variables and clauses, so a number of techniques have been described to narrow the number of variables [11,12,13,14,15,16,17].…”

Semantic Attachments for Domain-Independent Planning Systems

“…In order to generate an appropriate FDR description from PDDL/M tasks, we adapt the method of Helmert (Helmert 2009). Roughly, this process consists of the following phases: The generation of mutual exclusion (mutex) invariants that describe which propositions may never be true at the same time; a grounding phase in which, by means of a relaxed reachability analysis, a set of propositions (instantiations of predicates) is generated that may potentially be used in the planning process; the generation of a suitable FDR based on the mutex invariants and reachable propositions.…”