A New Approach for Tactical Decision Making in Lane Changing: Sample Efficient Deep Q Learning with a Safety Feedback Reward

@article{Yavas2020ANA,
  title={A New Approach for Tactical Decision Making in Lane Changing: Sample Efficient Deep Q Learning with a Safety Feedback Reward},
  author={M. U. Yavas and N. K. Ure and T. Kumbasar},
  journal={2020 IEEE Intelligent Vehicles Symposium (IV)},
  year={2020},
  pages={1156-1161}
}
  • M. U. Yavas, N. K. Ure, T. Kumbasar
  • Published 2020
  • Computer Science
  • 2020 IEEE Intelligent Vehicles Symposium (IV)
  • Automated lane change is one of the most challenging task to be solved of highly automated vehicles due to its safety-critical, uncertain and multi-agent nature. This paper presents the novel deployment of the state of art $\mathbf{Q}$ learning method, namely Rainbow DQN, that uses a new safety driven rewarding scheme to tackle the issues in an dynamic and uncertain simulation environment. We present various comparative results to show that our novel approach of having reward feedback from the… CONTINUE READING
    1 Citations

    Figures and Tables from this paper

    References

    SHOWING 1-10 OF 36 REFERENCES
    High-level Decision Making for Safe and Reasonable Autonomous Lane Changing using Reinforcement Learning
    • 33
    • PDF
    Automated Lane Change Decision Making using Deep Reinforcement Learning in Dynamic and Uncertain Highway Environment
    • 15
    • PDF
    Combining Planning and Deep Reinforcement Learning in Tactical Decision Making for Autonomous Driving
    • 27
    • PDF
    Automated Speed and Lane Change Decision Making using Deep Reinforcement Learning
    • C. Hoel, K. Wolff, L. Laine
    • Computer Science
    • 2018 21st International Conference on Intelligent Transportation Systems (ITSC)
    • 2018
    • 56
    • PDF
    Deep Reinforcement Learning: A Brief Survey
    • 595
    • PDF
    Reinforcement learning based overtaking decision-making for highway autonomous driving
    • Xin Li, X. Xu, Lei Zuo
    • Engineering
    • 2015 Sixth International Conference on Intelligent Control and Information Processing (ICICIP)
    • 2015
    • 27
    A machine learning approach for personalized autonomous lane change initiation and control
    • 38
    • PDF
    Deep reinforcement learning for robotic manipulation with asynchronous off-policy updates
    • 657
    • PDF
    Deep Reinforcement Learning with Double Q-Learning
    • 2,370
    • PDF