基于模型预测强化学习的内陆水道安全感知自主路径规划
摘要
近年来，由于将汽车与卡车移出市中心的趋势，城市水道中自主航运的兴趣显著增加。经典方法如基于 Frenet 坐标系的规划与势场导航常需调节众多配置参数，有时甚至需依情境采用不同配置。本文提出一种基于强化学习的新型路径规划方法，称为模型预测强化学习（MPRL）。MPRL 计算一系列供船舶跟随的航点。环境表示为占据栅格地图，使我们能处理任意形状的水道及任意数量与形状的障碍物。我们在两种场景中演示该方法，并将所得路径与基于 Frenet 坐标系的路径规划及基于近端策略优化（PPO）智能体的路径规划进行比较。结果表明 MPRL 在两种测试场景中均优于上述基线。基于 PPO 的方法在两种场景中均未能抵达目标，而 Frenet 坐标系方法在含障碍物的弯道场景中失败。MPRL 能在两种测试场景中安全（无碰撞）导航至目标。
引用
@article{arxiv.2311.09878,
title = {Safety Aware Autonomous Path Planning Using Model Predictive Reinforcement Learning for Inland Waterways},
author = {Astrid Vanneste and Simon Vanneste and Olivier Vasseur and Robin Janssens and Mattias Billast and Ali Anwar and Kevin Mets and Tom De Schepper and Siegfried Mercelis and Peter Hellinckx},
journal= {arXiv preprint arXiv:2311.09878},
year = {2023}
}
备注
\c{opyright} 2022 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works