Reinforcement Learning for Navigation of Mobile Robot with LiDAR

Inhwan Kim; Sarvar Hussain Nengroo; Dongsoo Har

doi:10.1109/ICECA52323.2021.9675949

Reinforcement Learning for Navigation of Mobile Robot with LiDAR

Inhwan Kim
, Sarvar Hussain Nengroo
, Dongsoo Har

Interdisciplinary Research Center for Sustainable Energy Systems

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

18 Scopus citations

Abstract

This paper presents a technique for navigation of mobile robot with Deep Q-Network (DQN) combined with Gated Recurrent Unit (GRU). The DQN integrated with the GRU allows action skipping for improved navigation performance. This technique aims at efficient navigation of mobile robot such as autonomous parking robot. Framework for reinforcement learning can be applied to the DQN combined with the GRU in a real environment, which can be modeled by the Partially Observable Markov Decision Process (POMDP). By allowing action skipping, the ability of the DQN combined with the GRU in learning key-action can be improved. The proposed algorithm is applied to explore the feasibility of solution in real environment by the ROS-Gazebo simulator, and the simulation results show that the proposed algorithm achieves improved performance in navigation and collision avoidance as compared to the results obtained by DQN alone and DQN combined with GRU without allowing action skipping.

Original language	English
Title of host publication	Proceedings of the 5th International Conference on Electronics, Communication and Aerospace Technology, ICECA 2021
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	148-154
Number of pages	7
ISBN (Electronic)	9781665435246
DOIs	https://doi.org/10.1109/ICECA52323.2021.9675949
State	Published - 2021

Publication series

Name	Proceedings of the 5th International Conference on Electronics, Communication and Aerospace Technology, ICECA 2021

Bibliographical note

Publisher Copyright:
© 2021 IEEE.

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

SDG 3 Good Health and Well-being

Keywords

Action skipping
Deep Q-Network
Gated Recurrent Unit
Navigation
Path planning
Reinforcement learning

ASJC Scopus subject areas

Fluid Flow and Transfer Processes
Artificial Intelligence
Computer Networks and Communications
Computer Science Applications
Aerospace Engineering
Electrical and Electronic Engineering
Industrial and Manufacturing Engineering

Access to Document

10.1109/ICECA52323.2021.9675949

Cite this

Kim, I., Nengroo, S. H., & Har, D. (2021). Reinforcement Learning for Navigation of Mobile Robot with LiDAR. In Proceedings of the 5th International Conference on Electronics, Communication and Aerospace Technology, ICECA 2021 (pp. 148-154). (Proceedings of the 5th International Conference on Electronics, Communication and Aerospace Technology, ICECA 2021). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/ICECA52323.2021.9675949

Kim, Inhwan ; Nengroo, Sarvar Hussain ; Har, Dongsoo. / Reinforcement Learning for Navigation of Mobile Robot with LiDAR. Proceedings of the 5th International Conference on Electronics, Communication and Aerospace Technology, ICECA 2021. Institute of Electrical and Electronics Engineers Inc., 2021. pp. 148-154 (Proceedings of the 5th International Conference on Electronics, Communication and Aerospace Technology, ICECA 2021).

@inproceedings{ae123a3416454990b5aba574c8772f22,

title = "Reinforcement Learning for Navigation of Mobile Robot with LiDAR",

abstract = "This paper presents a technique for navigation of mobile robot with Deep Q-Network (DQN) combined with Gated Recurrent Unit (GRU). The DQN integrated with the GRU allows action skipping for improved navigation performance. This technique aims at efficient navigation of mobile robot such as autonomous parking robot. Framework for reinforcement learning can be applied to the DQN combined with the GRU in a real environment, which can be modeled by the Partially Observable Markov Decision Process (POMDP). By allowing action skipping, the ability of the DQN combined with the GRU in learning key-action can be improved. The proposed algorithm is applied to explore the feasibility of solution in real environment by the ROS-Gazebo simulator, and the simulation results show that the proposed algorithm achieves improved performance in navigation and collision avoidance as compared to the results obtained by DQN alone and DQN combined with GRU without allowing action skipping.",

keywords = "Action skipping, Deep Q-Network, Gated Recurrent Unit, Navigation, Path planning, Reinforcement learning",

author = "Inhwan Kim and Nengroo, \{Sarvar Hussain\} and Dongsoo Har",

note = "Publisher Copyright: {\textcopyright} 2021 IEEE.",

year = "2021",

doi = "10.1109/ICECA52323.2021.9675949",

language = "English",

series = "Proceedings of the 5th International Conference on Electronics, Communication and Aerospace Technology, ICECA 2021",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

pages = "148--154",

booktitle = "Proceedings of the 5th International Conference on Electronics, Communication and Aerospace Technology, ICECA 2021",

address = "United States",

}

Kim, I, Nengroo, SH & Har, D 2021, Reinforcement Learning for Navigation of Mobile Robot with LiDAR. in Proceedings of the 5th International Conference on Electronics, Communication and Aerospace Technology, ICECA 2021. Proceedings of the 5th International Conference on Electronics, Communication and Aerospace Technology, ICECA 2021, Institute of Electrical and Electronics Engineers Inc., pp. 148-154. https://doi.org/10.1109/ICECA52323.2021.9675949

Reinforcement Learning for Navigation of Mobile Robot with LiDAR. / Kim, Inhwan; Nengroo, Sarvar Hussain; Har, Dongsoo.
Proceedings of the 5th International Conference on Electronics, Communication and Aerospace Technology, ICECA 2021. Institute of Electrical and Electronics Engineers Inc., 2021. p. 148-154 (Proceedings of the 5th International Conference on Electronics, Communication and Aerospace Technology, ICECA 2021).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Reinforcement Learning for Navigation of Mobile Robot with LiDAR

AU - Kim, Inhwan

AU - Nengroo, Sarvar Hussain

AU - Har, Dongsoo

PY - 2021

Y1 - 2021

N2 - This paper presents a technique for navigation of mobile robot with Deep Q-Network (DQN) combined with Gated Recurrent Unit (GRU). The DQN integrated with the GRU allows action skipping for improved navigation performance. This technique aims at efficient navigation of mobile robot such as autonomous parking robot. Framework for reinforcement learning can be applied to the DQN combined with the GRU in a real environment, which can be modeled by the Partially Observable Markov Decision Process (POMDP). By allowing action skipping, the ability of the DQN combined with the GRU in learning key-action can be improved. The proposed algorithm is applied to explore the feasibility of solution in real environment by the ROS-Gazebo simulator, and the simulation results show that the proposed algorithm achieves improved performance in navigation and collision avoidance as compared to the results obtained by DQN alone and DQN combined with GRU without allowing action skipping.

AB - This paper presents a technique for navigation of mobile robot with Deep Q-Network (DQN) combined with Gated Recurrent Unit (GRU). The DQN integrated with the GRU allows action skipping for improved navigation performance. This technique aims at efficient navigation of mobile robot such as autonomous parking robot. Framework for reinforcement learning can be applied to the DQN combined with the GRU in a real environment, which can be modeled by the Partially Observable Markov Decision Process (POMDP). By allowing action skipping, the ability of the DQN combined with the GRU in learning key-action can be improved. The proposed algorithm is applied to explore the feasibility of solution in real environment by the ROS-Gazebo simulator, and the simulation results show that the proposed algorithm achieves improved performance in navigation and collision avoidance as compared to the results obtained by DQN alone and DQN combined with GRU without allowing action skipping.

KW - Action skipping

KW - Deep Q-Network

KW - Gated Recurrent Unit

KW - Navigation

KW - Path planning

KW - Reinforcement learning

UR - https://www.scopus.com/pages/publications/85125404387

U2 - 10.1109/ICECA52323.2021.9675949

DO - 10.1109/ICECA52323.2021.9675949

M3 - Conference contribution

AN - SCOPUS:85125404387

T3 - Proceedings of the 5th International Conference on Electronics, Communication and Aerospace Technology, ICECA 2021

SP - 148

EP - 154

BT - Proceedings of the 5th International Conference on Electronics, Communication and Aerospace Technology, ICECA 2021

PB - Institute of Electrical and Electronics Engineers Inc.

ER -

Kim I, Nengroo SH, Har D. Reinforcement Learning for Navigation of Mobile Robot with LiDAR. In Proceedings of the 5th International Conference on Electronics, Communication and Aerospace Technology, ICECA 2021. Institute of Electrical and Electronics Engineers Inc. 2021. p. 148-154. (Proceedings of the 5th International Conference on Electronics, Communication and Aerospace Technology, ICECA 2021). doi: 10.1109/ICECA52323.2021.9675949

Reinforcement Learning for Navigation of Mobile Robot with LiDAR

Abstract

Publication series

Bibliographical note

UN SDGs

Keywords

ASJC Scopus subject areas

Access to Document

Fingerprint

Cite this