Effect-abstraction based relaxation for linear numeric planning

This paper studies an effect abstraction-based relaxation for reasoning about linear numeric planning problems. The effect abstraction decomposes non-constant linear numeric effects into actions with conditional, additive constant numeric effects. With little effort, on this abstracted version, it is possible to use known subgoaling-based relaxations and related heuristics. The combination of these two steps leads to a novel relaxation-based heuristic. Theoretically, the relaxation is proved tighter than the previous interval-based relaxation and leading to pruning-safe heuristics. Empirically, a heuristic developed on this relaxation leads to substantial improvements for a class of problems that are currently out of reach of state-of-the-art numeric planners.