首页|考虑社会性行为的自动驾驶运动规划研究综述

考虑社会性行为的自动驾驶运动规划研究综述

扫码查看
原文链接
  • NETL
  • NSTL
  • 万方数据
目前自动驾驶技术正处于从园区内示范运营向开放场景的大规模市场应用过渡的关键阶段.自动驾驶车辆若想应用于人们的日常生活,必将与人类驾驶的车辆在道路上长期共存,这势必会产生大量的交互场景,而如何恰当地与人类驾驶的车辆进行交互,则成为提升驾乘人员体验感及其对自动驾驶接受度的关键.人类的交互基于其社会属性,自动驾驶也应如此.但目前尚未有相关领域的完善综述.鉴于此,梳理其来龙去脉与最新进展.依次探讨传统方法所面临的驾驶困境、驾驶员社会性行为机理的相关研究和考虑社会属性的运动规划方法的最新进展与应用案例.在此基础上,总结现有方法的不足并对未来的研究方向进行展望.分析表明,结合社会学等领域的理论工具,分析驾驶员在复杂交通流中的社会性行为机理,并基于此构建符合大众预期的驾驶行为自动化决策体系,是未来自动驾驶社会性运动规划的主要研究方向.此外,进一步建立健全自动驾驶社会性行为的评价体系,同样对智能汽车的社会性人性化设计具有重要意义.
Review on Autonomous Vehicle Motion Planning Methods Considering Social Behavior
At present,autonomous driving technology is at a critical stage of transition from demonstration operations in parks to large-scale market applications in open scenarios.If self-driving vehicles are to be used in people's daily lives,they are bound to coexist with human-driven vehicles on the road for a long time,which inevitably generates a large number of interaction scenarios,and how to properly interact with human-driven vehicles becomes the key to improving the driver's experience and their acceptance of autonomous driving.Human interaction is based on its social attributes,and so should autonomous vehicles.However,there is no comprehensive review in this area.In light of this,the ins and outs and recent developments are reviewed.The driving dilemma faced by traditional methods,the research on the mechanism of driver social behaviour,and the recent progress and application cases of motion planning methods considering social attributes are discussed in turn.On this basis,the shortcomings of the existing methods are summarized and future research directions are prospected.The analysis shows that combining theoretical tools from sociology and other fields,analysing drivers'social behavioural mechanisms in complex traffic flows,and constructing an automated decision-making system for driving behaviour that meets public expectations based on this is the main research direction for future social motion planning for autonomous driving.In addition,further establis0hing a sound evaluation system for the social behaviour of autonomous driving is equally important for the social humanization design of intelligent vehicles.

autonomous vehiclessocial compliancemotion planninginteraction behaviournew mixed traffic flow

高镇海、于桐、孙天骏

展开 >

吉林大学汽车工程学院 长春 130022

吉林大学汽车底盘集成与仿生全国重点实验室 长春 130022

自动驾驶 社会属性 运动规划 交互行为 混合交通流

教育部产学合作育人吉林大学研究生创新基金吉林大学长沙汽车创新研究院自由探索中央高校基本科研业务费专项资金汽车动力传动与电子控制湖北省重点实验室开放基金

231007538181742451230411061CAIRIZT202201062022-JCXK-24ZDK12023A05

2024

10.3901/JME.2024.10.112

机械工程学报
中国机械工程学会

机械工程学报

CSTPCD北大核心
影响因子：1.362
ISSN：0577-6686
年,卷(期)：2024.60(10)