Shravan S Rai*

Arizona State University, USA

*Corresponding Author: Shravan S Rai, Arizona State University, USA.

Received: April 17, 2024; Published: April 30, 2024

Abstract

This paper presents an innovative approach to obstacle avoidance in autonomous systems through the application of reinforcement learning techniques. This comprehensive study delves into the intricate challenges of navigating autonomous systems in dynamic environments. By harnessing the power of advanced reinforcement learning techniques, specifically a synergistic blend of Q-learning and the Deep Deterministic Policy Gradient (DDPG) algorithm, this effort has pioneered a groundbreaking approach to improve obstacle avoidance capabilities of a mobile robot. This research meticulously explores the implementation and efficacy of this model through rigorous testing in a variety of simulated scenarios. The results obtained are not only promising but signify a substantial leap forward from traditional methods. There are marked enhancements observed in both the efficiency and safety aspects of autonomous navigation, paving the way for more sophisticated and resilient obstacle avoidance strategies. This investigation not only makes a significant contribution to the evolving field of robotics and artificial intelligence but also lays the groundwork for future explorations into the potential applications of reinforcement learning in complex, ever-changing environments. The findings in this paper offer valuable insights and a solid foundation for subsequent research aimed at optimizing autonomous systems for enhanced operational performance and adaptability.

Keywords: Obstacle; Avoidance; Reinforcement; Learning; DDPG

References

1. G Grisetti., et al. “A fully autonomous indoor quadrotor”. in: IEEE Transactions on Robotics 28 (2012): 90-100.
2. MW Mueller and R D’Andrea. “A model predictive controller for quadrocopter state interception”. In: Control Conference (ECC), 2013 European (2013): 1383-1389.
3. D Silver., et al. “Playing atari with deep reinforcement learning”. in: arXiv preprint arXiv:1312.5602, 2013. (2013).
4. Richard S Sutton and Shangtong Zhang. “A Deeper Look at Experience Replay”. In: NIPS 2017 Deep Reinforcement Learning Symposium (2017).
5. Stephen James., et al. “PyRep: Bringing V-REP to Deep Robot Learning”. In: arXiv preprint arXiv:1906.11176 (2019).
6. Alexander Pritzel., et al. “Continuous control with deep reinforcement learning”. In: Google Deepmind (2019).
7. Jiahao Lu., et al. “Efficient hindsight reinforcement learning using demonstrations for robotic tasks with sparse rewards”. In: International Journal of Advanced Robotic Systems (2020).
8. Ankit Choudhary. “A Hands-On Introduction to Deep Q-Learning using OpenAI Gym in Python”.
9. Or Rivlin. “Reinforcement Learning with Hindsight Experience Replay”.
10. Open-AI Gym extension for robotics based on V-REP.
11. Q-learning.

Citation

Citation: Shravan S Rai. “Autonomous Mobile Robot Obstacle Avoidance with Reinforcement Learning".Acta Scientific Computer Sciences 6.5 (2024): 16-21.

Copyright

Copyright: © 2024 Shravan S Rai. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.