Mobile Robot Control Awareness System Using Deep Reinforcement Learning Technique Based on DQN and DDPG Algorithm
Pham Van Huy, Pham Van Minh, Nguyen Cao Cuong, Le Van Anh, Tran Duc Chuyen*
Faculty of Electrical Engineering, University of Economics – Technology for Industries, No. 353 Tran Hung Dao, Nam Dinh, VietNam.
Corresponding Author Email: [email protected]
This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
The mobile robots can perform a variety of real-world tasks such as environmental monitoring, delivery, search and rescue. Such tasks require varying degrees of self direction in response to environmental changes. However, most of the navigation methods rely on static obstacle maps and are not self taught. In this paper, we propose to use two deep reinforcement learning techniques: Deep Q Network (DQN) and Deep Deterministic Policy Gradient (DDPG) to navigate the mobile robot in an indeterminate flat environment. The simulation results on the Gazebo software show the feasibility of the proposed method. Robots can safely complete navigation tasks in an unprotected dynamic environment and become a truly intelligent system with strong self-learning and adaptability.