Abstract
The reinforcement learning algorithms have been proven to be extremely accurate in performing a variety of tasks. These algorithms have outperformed humans in traditional games. This paper proposes a reinforcement learning based approach to autonomous driving. The autonomous vehicles must be able to deal with all external situations to ensure safety and to avoid undesired circumstances such as collisions. Thus, we propose the use of deep deterministic policy gradient (DDPG) algorithm which is able to work in a complex and continuous domain. To avoid physical damage and reduce costs, we choose to use a simulator to test the proposed approach. The CARLA simulator would be used as the environment. To fit the DDPG algorithm to the CARLA environment, our network architecture consists of critic and actor networks. The performance would be evaluated based on rewards generated by the agent while driving in the simulated environment.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Similar content being viewed by others
References
S. Singh, Critical reasons for crashes investigated in the national motor vehicle crash causation survey. Technical Report, Traffic Safety Facts Crash Stats, National Highway Traffic Safety Administration (2015)
V. Mnith, K. Kavukcuoglu, D. Silver, A. Graves, L. Antonoglou, D. Wierstra, M. Riedmiller, Playing Atari with deep reinforcement learning. Tech. Rep. (2013). https://arxiv.org/abs/1312.5602
H. Daume, A course in machine learning (2017). http://ciml.info/dl/v0_99/ciml-v0_99-ch18.pdf
F. Codevilla, M. Muller, A. Lopez, V. Koltun, A. Dosovitskiy, End-to-end driving via conditional imitation learning, in IEEE International Conference on Robotics and Automation (ICRA) (IEEE Press, Brisbane, QLD, 2018), pp. 4693–4700
T. Osa, J. Pajarinen, G. Neumann, J.A. Bagnell, P. Abbeel, J. Peters, An algorithm perspective on imitation learning. Found. Trends Robot. 7(1–2), 1–179 (2018)
A. Dosovitskiy, G. Ros, F. Codevilla, A. Lopez, V. Koltun, CARLA: an open driving simulator, in 1st Conference of Robot Learning (CoRL 2017) (PMLR Press, Mountain View, USA, 2017), pp. 1–16
Y. Pan, C. Cheng, K. Saigol, K. Lee, X. Yan, E. Theodorou, B. Boots, Agile autonomous driving using end-to-end deep imitation learning. Tech. Rep. (2019). http://arxiv.org/abs/1709.07174
Open AI Spinning Up. https://spinningup.openai.com/en/latest/algorithms/ddpg.html. Accessed 10 Jan 2020
H. Yi, Deep deterministic policy gradient for autonomous vehicle driving, in 20th International Conference on Artificial Intelligence (ICAI’18) (CSREA Press, Nevada, USA, 2018), pp. 191–194
M. Bojarski, D. Del Testa, D. Dworakowski, B. Firner, B. Flepp, P. Goyal, L. Jackel, M. Monfort, U. Muller, J. Zhang, X. Zhang, J. Zhao, K. Zeiba, End to end learning for self driving cars. Tech. Rep. (2016). http://arxiv.org/abs/1604.07316
F. Chollet, Deep Learning with Python (Manning, New York, 2018)
E. Bibbona, G. Panfilo, P. Tavella, The Ornstein-Uhlenbeck process as a model of a low pass filtered white noise. J. Metrol. 45, 117–126 (2008)
A. Sauer, N. Savinov, A. Geiger, Conditional affordance learning for driving in urban environments, in 2nd Conference on Robot Learning (CoRL 2018) (PMLR Press, Zurich, Switzerland, 2018), pp. 237–252
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2021 The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.
About this paper
Cite this paper
Kadam, R., Vidhani, V., Valecha, B., Bane, A., Giri, N. (2021). Autonomous Vehicle Simulation Using Deep Reinforcement Learning. In: Joshi, A., Khosravy, M., Gupta, N. (eds) Machine Learning for Predictive Analysis. Lecture Notes in Networks and Systems, vol 141. Springer, Singapore. https://doi.org/10.1007/978-981-15-7106-0_53
Download citation
DOI: https://doi.org/10.1007/978-981-15-7106-0_53
Published:
Publisher Name: Springer, Singapore
Print ISBN: 978-981-15-7105-3
Online ISBN: 978-981-15-7106-0
eBook Packages: Intelligent Technologies and RoboticsIntelligent Technologies and Robotics (R0)