基于强化学习经验优先提取的汽车纵向多态控制
Longitudinal polymorphic control of vehicle based on reinforcement learning with prioritized experience extraction
黄鹤 1付梦园 1吴润晨 1黄泽辰 1曾琦 1石琴1
作者信息
- 1. 合肥工业大学汽车与交通工程学院,安徽合肥 230009;合肥工业大学智能制造技术研究院,安徽合肥 230009;安徽省智慧交通车路协同工程研究中心,安徽合肥 230009
- 折叠
摘要
文章提出一种引入经验优先提取(prioritized experience extraction,PEE)规则的深度Q网络(deep Q network,DQN)算法,用于解决汽车纵向行驶时的多态控制问题.首先,建立车辆纵向力矩传递模型和强化学习算法模型,在进行算法移植以及制定奖励函数时综合考虑车速、距离等相关因素的综合限制;然后,通过仿真与硬件在环实验验证强化学习算法在汽车纵向多态控制方面的有效性;最后,引入PEE规则提高常规DQN算法的计算效率,解决算法区域性过拟合问题.PEE规则的引入有助于平滑主车的跟随车速,与相对距离相配合提升了行驶时的舒适性与安全性.
Abstract
In order to solve the polymorphic control problem of the longitudinal motion of vehicle,a deep Q network(DQN)algorithm based on the rule of prioritized experience extraction(PEE)was put forward.The vehicle longitudinal torque transfer model and reinforcement learning algorithm model were analyzed and established.The transplantation of the algorithm and the formulation of reward function were made taking the comprehensive limitations of the relevant factors such as speed and dis-tance into consideration.Through the simulation and hardware-in-the-loop experiment,the effective-ness of deep reinforcement learning algorithm in the longitudinal polymorphic control of vehicle is ver-ified.In addition,PEE rule was introduced to improve the computational efficiency of conventional DQN algorithm and solve the overfitting problem to some extent.The PEE rule also realizes the smooth following speed of the main vehicle,which,in combination with the relative distance,im-proves the comfort and safety during driving.
关键词
深度强化学习/纵向控制/多态控制/经验优先提取(PEE)规则Key words
deep reinforcement learning/longitudinal control/polymorphic control/prioritized experi-ence extraction(PEE)rule引用本文复制引用
基金项目
国家自然科学基金(71971073)
长三角科技创新共同体联合攻关资助项目(2023)(2023CSJGG1600)
安徽省新能源汽车暨智能网联汽车创新工程项目(GXXT-2020-076)
合工大智能院"科技成果转化及产业化"资助项目(2019)(IMICZ2019005)
出版年
2024
NETL
NSTL
万方数据