Polynomial-Time in PDDL Input Size:Making the Delete Relaxation Feasible for Lifted Planning
| Autor: | Lauer, Pascal, Torralba, Alvaro, Fišer, Daniel, Höller, Daniel, Wichlacz, Julia, Hoffmann, Jörg |
|---|---|
| Jazyk: | angličtina |
| Rok vydání: | 2021 |
| Zdroj: | Lauer, P, Torralba, A, Fišer, D, Höller, D, Wichlacz, J & Hoffmann, J 2021, Polynomial-Time in PDDL Input Size : Making the Delete Relaxation Feasible for Lifted Planning . in Proceedings of the Thirtieth International Joint Conference on Artificial Intelligence. IJCAI-21 . International Joint Conferences on Artificial Intelligence, pp. 4119-4126, Thirtieth International Joint Conference on Artificial Intelligence. IJCAI-21, Canada, 19/08/2021 . < https://www.ijcai.org/proceedings/2021/0567.pdf > |
| Popis: | Polynomial-time heuristic functions for planningare commonplace since 20 years. But polynomialtime in which input? Almost all existing approaches are based on a grounded task representation, not on the actual PDDL input which is exponentially smaller. This limits practical applicability to cases where the grounded representationis “small enough”. Previous attempts to tackle thisproblem for the delete relaxation leveraged symmetries to reduce the blow-up. Here we take a moreradical approach, applying an additional relaxationto obtain a heuristic function that runs in time polynomial in the size of the PDDL input. Our relaxation splits the predicates into smaller predicatesof fixed arity K. We show that computing a relaxed plan is still NP-hard (in PDDL input size) forK ≥ 2, but is polynomial-time for K = 1. We implement a heuristic function for K = 1 and showthat it can improve the state of the art on benchmarks whose grounded representation is large. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|