Combine Incremental Planning, Reasonable Ordering, External Information and Soft Goals Handling

Partial satisfaction or Over-subscription Planning problems arise in many real world applications, and these soft problem goals have been defined as preferences in PDDL3.0. Planning with over-subscription problem and preferences is one of the most challenging problems in planning. Inspired by the following two facts, we come up with a method to handle soft goals problem, and it also efficient with classical planning problems. Firstly, using external information has turned out to be a fruitful approach for classical planning; secondly, strategies in incremental planning for ordering and grouping subproblems partitioned by the subgoals of a planning problem have recently draw a lot attention from planning community. In this paper we present an approach to combining incremental planning with control knowledge and this approach can be used to handle soft goals. Our method improved incremental planning using goal agenda with more ordering for atomic goal and grouping subproblems partitioned by the subgoals. We also provide preliminary empirical results in a planner called FFps which demonstrate the effectiveness of our approach in comparison to the original FF and it can handle soft goals problem.