首页|强化学习在自动驾驶换道研究中的应用

强化学习在自动驾驶换道研究中的应用

扫码查看
在车辆行驶的过程中,车辆的硬件设备有保障的前提下,错误的行为决策是交通事故发生的主要诱因.强化学习能够和环境不断进行交互调整,具有很好的泛化能力,适合复杂的换道决策场景.首先对自动驾驶系统和换道决策进行了简要介绍;其次,介绍了强化学习的原理和其具代表性算法,并总结了强化学习在换道决策系统中的应用;最后,根据存在的问题,对强化学习在自动驾驶换道决策中的应用进行了展望.
Application of Reinforcement Learning to Study of Autonomous Driving for Lane Changing
In the process of vehicle driving,the vehicle's hardware equipment is guaranteed under the premise that the wrong behavioral decision is the main causative factor of traffic accidents.Reinforcement learning is able to continuously interact and adjust with the environment,has good generalization ability,and is suitable for com-plex lane-changing decision-making scenarios.Firstly,a brief introduction to the automatic driving system and lane-changing decision-making is given;secondly,the principle of reinforcement learning and its representative algorithms are introduced,and the application of reinforcement learning in lane-changing decision-making system is summarized;finally,based on the existing problems,an outlook is given to the application of reinforcement learning in automatic driving lane-changing decision-making.

Reinforcement learningAutomatic drivingLane changing

杜婉、董天悦、罗玉玲

展开 >

西安汽车职业大学,陕西西安 710600

强化学习 自动驾驶 换道

2024

内燃机与配件
石家庄金刚内燃机零部件集团有限公司

内燃机与配件

影响因子:0.095
ISSN:1674-957X
年,卷(期):2024.(12)