cs.AI updates on arXiv.org 15小时前
Reinforced Imitative Trajectory Planning for Urban Automated Driving
index_new5.html
../../../zaker_core/zaker_tpl_static/wap/tpl_guoji1.html

 

本文提出一种结合强化学习和模仿学习的新方法,用于解决城市自动驾驶中的轨迹规划问题,通过开发基于transformer的贝叶斯奖励函数,实现多步规划,并在nuPlan数据集上验证了其优越性。

arXiv:2410.15607v2 Announce Type: replace-cross Abstract: Reinforcement learning (RL) faces challenges in trajectory planning for urban automated driving due to the poor convergence of RL and the difficulty in designing reward functions. Consequently, few RL-based trajectory planning methods can achieve performance comparable to that of imitation learning-based methods. The convergence problem is alleviated by combining RL with supervised learning. However, most existing approaches only reason one step ahead and lack the capability to plan for multiple future steps. Besides, although inverse reinforcement learning holds promise for solving the reward function design issue, existing methods for automated driving impose a linear structure assumption on reward functions, making them difficult to apply to urban automated driving. In light of these challenges, this paper proposes a novel RL-based trajectory planning method that integrates RL with imitation learning to enable multi-step planning. Furthermore, a transformer-based Bayesian reward function is developed, providing effective reward signals for RL in urban scenarios. Moreover, a hybrid-driven trajectory planning framework is proposed to enhance safety and interpretability. The proposed methods were validated on the large-scale real-world urban automated driving nuPlan dataset. Evaluated using closed-loop metrics, the results demonstrated that the proposed method significantly outperformed the baseline employing the identical policy model structure and achieved competitive performance compared to the state-of-the-art method. The code is available at https://github.com/Zigned/nuplan_zigned.

Fish AI Reader

Fish AI Reader

AI辅助创作,多种专业模板,深度分析,高质量内容生成。从观点提取到深度思考,FishAI为您提供全方位的创作支持。新版本引入自定义参数,让您的创作更加个性化和精准。

FishAI

FishAI

鱼阅,AI 时代的下一个智能信息助手,助你摆脱信息焦虑

联系邮箱 441953276@qq.com

相关标签

强化学习 自动驾驶 轨迹规划 模仿学习 贝叶斯奖励函数
相关文章