用于陆地移动机器人无地图导航的并行分布深度强化学习
机器人学
2024-09-04 v2
摘要
本文介绍了利用并行分布actor-critic网络进行陆地移动机器人导航的新型深度强化学习（Deep-RL）技术。我们的方法使用激光测距、相对距离和目标角度来引导机器人。我们在Gazebo仿真器中训练智能体并在真实场景中部署。结果表明，并行分布Deep-RL算法增强了决策能力，并在导航和空间泛化方面优于非分布和行为基方法。
引用
@article{arxiv.2408.05744,
title = {Parallel Distributional Deep Reinforcement Learning for Mapless Navigation of Terrestrial Mobile Robots},
author = {Victor Augusto Kich and Alisson Henrique Kolling and Junior Costa de Jesus and Gabriel V. Heisler and Hiago Jacobs and Jair Augusto Bottega and André L. da S. Kelbouscas and Akihisa Ohya and Ricardo Bedin Grando and Paulo Lilles Jorge Drews-Jr and Daniel Fernando Tello Gamarra},
journal= {arXiv preprint arXiv:2408.05744},
year = {2024}
}
备注
Paper accepted at the 24th International Conference on Control, Automation and Systems (ICCAS)