| 面向智能车辆的现役道路设施行驶适应性研究综述 |
| |
| 引用本文： | 于斌,王书易,马羊,杨倩,周雯,刘晋周.面向智能车辆的现役道路设施行驶适应性研究综述[J].中国公路学报,2022,35(10):205-225. |
| |
| 作者姓名： | 于斌 王书易 马羊 杨倩 周雯 刘晋周 |
| |
| 作者单位： | 1. 东南大学 交通学院, 江苏 南京 211189;2. 合肥工业大学 汽车与交通工程学院, 安徽 合肥 230009 |
| |
| 基金项目： | 国家自然科学基金项目(51878163)；江苏省研究生科研创新基金项目(KYCX18_0152)； 江苏省交通运输科技与成果转化项目(2020Y19-1(1)) |
| |
| 摘 要： | 为了总结面向智能车辆的现役道路设施行驶适应性，即现役道路基础设施承载智能车辆行驶的适宜程度，阐述自主智能驾驶定义与驾驶自动化等级分类，在此基础上剖析不同等级间的人机功能差异，并分别从感知层、感知-决策层、决策-控制层探讨与道路设计要素相关联的人机功能差异，通过归纳总结智能车辆与道路几何要素、路面性能及其他道路要素(如道路标线)的相互作用机制研究，从道路工程角度及其他道路要素方面回顾该领域的研究现状，指出存在的问题和未来发展方向。研究结果表明：相比传统车辆，配置高等级自动驾驶系统的智能车辆对现役道路设施行驶适应性最高，主动安全系统次之，而驾驶辅助及有条件自动驾驶系统适应性不足。而目前研究主要问题包括：难以归纳、标定不同驾驶自动化等级间的人机功能差异及其对于道路设计参数的需求设计值；测试道路场景条件过于理想，考虑的驾驶自动化等级单一，试验规模和样本有限；道路几何、路面性能以及道路标志、标线等道路要素与智能车辆间的相互作用机制研究不足，缺乏与不同道路场景相匹配的智能车辆驾驶特征数据的获取手段。因此建议：重视并推动与道路设计要素相关联的关键人机功能差异指标信息共享；联合高保真且可交互的道路场景、高精度感知传感器物理模型、车辆动力学模型及微观交通流模型，利用测试场景自动化生成、极限工况场景搜寻与泛化等技术开展智能驾驶虚拟测试，突破现有研究的深度和广度；探索反映不同等级智能车辆的道路行驶适应性特征指标与评价标准，精准、有效地评估预测复杂道路场景及不利道路条件下的行驶适应性。
|
| 关 键 词： | 交通工程 智能车辆 综述 道路行驶适应性 驾驶自动化等级 道路几何 道路路面 |
| 收稿时间： | 2021-10-25 |
Review of Driving Adaptability of Intelligent Vehicles to As-built Roadway Infrastructures |
| |
| YU Bin,WANG Shu-yi,MA Yang,YANG Qian,ZHOU Wen,LIU Jin-zhou.Review of Driving Adaptability of Intelligent Vehicles to As-built Roadway Infrastructures[J].China Journal of Highway and Transport,2022,35(10):205-225. |
| |
| Authors: | YU Bin WANG Shu-yi MA Yang YANG Qian ZHOU Wen LIU Jin-zhou |
| |
| Affiliation: | 1. School of Transportation, Southeast University, Nanjing 211189, Jiangsu, China;2. School of Automotive and Transportation Engineering, Hefei University of Technology, Hefei 230009, Anhui, China |
| |
| Abstract: | To summarize the driving adaptability of intelligent vehicles to as-built roadway infrastructures, i.e., the readiness of as-built roadway infrastructures for intelligent vehicles, the definition of disconnected intelligent vehicles and the classification of driving automation levels were clarified. On this basis, the differences in human versus (vs.) machine functions between varied levels were analyzed. Then, those differences in human vs. machine functions associated with roadway design elements were explored from three layers: perception layer, perception-planning layer, and planning-control layer. By summarizing studies on the interaction mechanism between intelligent vehicles and road geometric elements, pavement performance, and other road-related elements (e.g., road markings), related work was reviewed from the perspective of road engineering and other road-related elements, followed by the proposed limitations and future work. Results show that compared with the human-driven traditional vehicle, the 'Automated Driving System-dedicated’ vehicle enhances the adaptability to as-built roads with the most potential, followed by the vehicle equipped with active safety systems, while the driver assistance and partial/conditional driving automation are inadequate in adaptability. The main problems of existing studies include: It is difficult to summarize and calibrate those functional differences between varied levels and their required design values for road design parameters; Testing road scenarios are mostly in ideal conditions. The considered automation levels are limited. Experiments are limited in the scale and considered samples; The interaction mechanisms between intelligent vehicles and road geometric elements, pavement performance, and road signs & markings condition or other road-related elements were studied insufficiently. Researchers lack approaches to obtain data on the driving characteristics of intelligent vehicles corresponding to different driving scenarios. Therefore, it is suggested that: The information-sharing of critical parameters of human vs. machine functional differences associated with roadway design elements should be emphasized and promoted; Combining realistic and interactive road scenarios, high-precision physical perception sensor models, vehicle dynamics models, and microscopic traffic flow models, and utilizing technologies of the automatic generation of virtual test scenarios, and search and generalization of corner-case scenarios to conduct virtual simulations for intelligent vehicles to break through the depth and breadth of research in this field. Explore the characteristic indexes and evaluation criteria of driving adaptability of different-levels intelligent vehicles to roads, and then evaluate and predict that adaptability under complicated road scenarios and adverse roadway conditions more accurately and effectively. |
| |
| Keywords: | traffic engineering intelligent vehicle review driving adaptability to roadway infrastructures driving automation level road geometry road pavement |
|
| 点击此处可从《中国公路学报》浏览原始摘要信息 |
|
点击此处可从《中国公路学报》下载全文 |
|
