Skip to content

Improved Robustness and Safety for Autonomous Vehicle Control with Adversarial Reinforcement Learning #29

Description

@DeepTecher

Improved Robustness and Safety for Autonomous Vehicle Control with Adversarial Reinforcement Learning

提交日期; 2019-03-08
团队: 斯坦福大学航空航天系
作者: Xiaobai Ma, Katherine Driggs-Campbell, Mykel J. Kochenderfer

摘要:为了对自动驾驶车提高效率并减少故障,研究重点是开发考虑到环境干扰的强大而安全的学习方法。 强化学习现有文献将学习问题作为自治系统与干扰之间的双人游戏。 本文研究了两种不同的算法来解决这种双人游戏:鲁棒性对抗强化学习和神经虚拟自我游戏,并比较这两种算法在自动驾驶场景的表现。我们将游戏制定扩展到半竞争环境,并证明最终的对手可以更好地捕获有意义的干扰,从而提高整体性能。 与传统强化学习方法产生的基线控制政策相比,这种方式下的鲁棒性的政策表现提高了驾驶效率,同时降低了碰撞率。

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions