Abstract:
Knowledge based approaches developed for AI planning can convert an intractable planning problem to a tractable one. Current techniques often use temporal logics to expre...Show MoreMetadata
Abstract:
Knowledge based approaches developed for AI planning can convert an intractable planning problem to a tractable one. Current techniques often use temporal logics to express Search Control Knowledge (SCK) in logic based planning. However, traditional temporal logics are limited in expressiveness since they are unable to express spatial constraints which are as important as temporal ones in many planning domains. To this end, we propose a two-dimensional (spatial and temporal) logic namely PPTLSL by temporalizing separation logic with PPTL (Propositional Projection Temporal Logic) which is well-suited to specify SCK involving both spatial and temporal constraints in planning. We prove that PPTLSL is decidable essentially via an equisatisfiable translation from PPTLSL to its restricted form. Moreover, we implement a tool, S-TSolver, which effectively computes plans under the guidance of the spatio-temporal SCK expressed by PPTLSL formulas. The effectiveness of the tool is evaluated on selected benchmark domains from the International Planning Competition.
Published in: IEEE Transactions on Knowledge and Data Engineering ( Volume: 30, Issue: 10, 01 October 2018)