Abstract
This work presents a theory of planning called Case-Based Planning that suggests the use of a planner's dynamic memory of experience rather than a static base of rules. This view of planning from experience is supported by a learning system that incorporates new experiences into the planner's episodic memory. A process of plan repair is also presented in which plan failures are diagnosed through a casual analysis of the steps and states that led to their occurrence. Successful plans, failures, and past repairs are stored, with successes being accessed and modified to create new plans, failures being used to warn the planner of impending problems, and repair strategies being referenced to tell the planner how to deal with faulty plans at hand. By storing failures as well as successes, the planner is able to anticipate and avoid future plan failures. These ideas of memory, learning, and planning are implemented in the case-based planner CHEF, which creates new plans in the domain of Szechwan cooking.
Index Terms
- New Books
Recommendations
New results about LCGP, a least committed GraphPlan
AIPS'00: Proceedings of the Fifth International Conference on Artificial Intelligence Planning SystemsPlanners from the family of Graphplan (Graphplan, IPP, STAN...) are presently considered as the most efficient ones on numerous planning domains. Their partially ordered plans can be represented as sequences of sets of simultaneous actions. Using this ...
New advances in GraphHTN: identifying independent subproblems in large HTN domains
AIPS'00: Proceedings of the Fifth International Conference on Artificial Intelligence Planning SystemsWe describe in this paper a new method for extracting knowledge on Hierarchical Task-Network (HTN) planning problems for speeding up the search. This knowledge is gathered by propagating properties through an AND/OR tree that represents disjunctively ...
Comments