作者: Sebastian Thrun , Joelle Pineau , Geoff Gordon
DOI:
关键词:
摘要: This paper presents a scalable control algorithm that enables deployed mobile robot to make high-level decisions under full consideration of its probabilistic belief. We draw on insights from the rich literature structured controllers and hierarchical MDPs propose PolCA, which learns both subtask-specific state abstractions policies. The resulting controller has been successfully implemented onboard robotic assistant in nursing facility. To best our knowledge, this work is unique instance applying POMDPs highlevel problems.