Real-time hierarchical POMDPs for autonomous robot navigation |
| |
| Affiliation: | 1. Kazan Institute of Biochemistry and Biophysics Kazan Scientific Center of the Russian Academy of Sciences, Post Box 30, 420111 Kazan, Russia;2. Institute of Physiology, Academy of Sciences of the Czech Republic, Videnska 1083, 14200 Prague, Czech Republic;3. Department of Physiology, Faculty of Sciences, Charles University, Vinicna 7, Prague, Czech Republic;4. Kazan State Medical University, Butlerov st. 49, 420012 Kazan, Russia;5. Kazan Federal University, Kremlyovskaya st. 18, 420008 Kazan, Russia;1. School of Chemical Engineering, National Technical University of Athens, Athens, Greece;2. Department of Civil and Environmental Engineering, University of Cyprus, Nicosia, Cyprus;3. Titan Cement Company S.A., Group R&D and Quality Dept., Kamari Plant, Elefsina, Greece;1. Institute of Oceanology, Shanghai Jiao Tong University, Shanghai 200240, China;2. Centre for Maritime Engineering, Control and Imaging, School of Computer Science, Engineering and Mathematics, Flinders University, SA, Australia |
| |
| Abstract: | This paper proposes a new hierarchical formulation of POMDPs for autonomous robot navigation that can be solved in real-time, and is memory efficient. It will be referred to in this paper as the Robot Navigation–Hierarchical POMDP (RN-HPOMDP). The RN-HPOMDP is utilized as a unified framework for autonomous robot navigation in dynamic environments. As such, it is used for localization, planning and local obstacle avoidance. Hence, the RN-HPOMDP decides at each time step the actions the robot should execute, without the intervention of any other external module for obstacle avoidance or localization. Our approach employs state space and action space hierarchy, and can effectively model large environments at a fine resolution. Finally, the notion of the reference POMDP is introduced. The latter holds all the information regarding motion and sensor uncertainty, which makes the proposed hierarchical structure memory efficient and enables fast learning. The RN-HPOMDP has been experimentally validated in real dynamic environments. |
| |
| Keywords: | |
| 本文献已被 ScienceDirect 等数据库收录！ |
|
