| 人机混驾环境下混行车辆雾模型研究 |
| |
| 引用本文： | 梁军,徐鹏,蔡英凤,陈龙,刘擎超.人机混驾环境下混行车辆雾模型研究[J].中国公路学报,2021,34(11):255-264. |
| |
| 作者姓名： | 梁军 徐鹏 蔡英凤 陈龙 刘擎超 |
| |
| 作者单位： | 江苏大学 汽车工程研究院, 江苏 镇江 212013 |
| |
| 基金项目： | 国家重点研发计划项目（2018YFB1600500）；国家自然科学基金项目（51875255）；江苏省六大人才高峰项目（2018-TD-GDZB-022） |
| |
| 摘 要： | 目前搭载高级驾驶辅助系统和车联网（Vehicular Ad Hoc Network，VANET）技术的智能网联车（Intelligent Connected Vehicles，ICV）正大量涌入人工驾驶车（Manual Vehicle，MV）流之中，ICV与MV共存的异构车辆混行交通态势逐步形成，异构车辆之间的交互产生壁垒。混行之下单个ICV虽可依托单车硬件传感与单车计算单元实现与MV的交互意图识别，但其受有限算力与有限传感的影响，资源负载增大，时效性与安全性方面存在一定的误差与风险，而混行之下的VANET技术也不能够提供全局性车路资源用以高度匹配ICV与MV的交互场景，而且越来越多的ICV计算需求也在激增VANET的负载压力。对此，结合边缘计算概念中的雾计算理论，提出混行车辆雾模型（Mixed Vehicle Fog，MVF），充分发挥车联网络边缘节点能力，通过合理整合调度ICV资源的方法，解决对MV正常交互意图计算的时效性与安全性问题。该模型首先通过各感知单元响应混行交通环境下ICV与MV的正常交互事件，然后利用基于容错节点分簇的资源调度算法（Fault-tolerant Node Clustering Resource Scheduling Algorithm，FNC-RSA），动态划分局部路段内对交互事件具有相关意图感知与计算需求的ICV为一组"协同雾群"，再评估雾内ICV节点自身资源与路由代价，定向定量调度资源，最终实现雾群内部MV交互信息共享与驾驶意图协同计算。试验借助Prescan和MATLAB搭建联合仿真平台，与低能耗自适应分簇型路由算法（Low Energy Adaptive Clustering Hierarchy，LEACH）模型对比，验证MVF模型的运行效率与模型鲁棒性。研究结果表明：MVF模型通过交互事件细分协同雾群，保证了计算负载均衡，提高了ICV定向资源计算与传输效率，比LEACH模型降低了55.17%的平均跳数，缩短了45.40%的平均任务完成时间，抗时延干扰能力强，鲁棒性能优异。该模型对于打破混行环境异构车辆交互壁垒，提高混行道路交通行车安全，创造车联网络良性发展空间具有积极作用。
|
| 关 键 词： | 交通工程 MVF模型 FNC-RSA算法 智能网联车 混行交通 |
| 收稿时间： | 2020-02-20 |
Research on Mixed Vehicle Fog Model and Key Algorithm in Human-machine Hybrid Driving Environment |
| |
| LIANG Jun,XU Peng,CAI Ying-feng,CHEN Long,LIU Qing-chao.Research on Mixed Vehicle Fog Model and Key Algorithm in Human-machine Hybrid Driving Environment[J].China Journal of Highway and Transport,2021,34(11):255-264. |
| |
| Authors: | LIANG Jun XU Peng CAI Ying-feng CHEN Long LIU Qing-chao |
| |
| Affiliation: | Automotive Engineering Research Institute, Jiangsu University, Zhenjiang 212013, Jiangsu, China |
| |
| Abstract: | Various intelligent connected vehicles (ICVs) equipped with advanced driving assistance systems and vehicular ad hoc networks (VANETs) are being incorporated into the flow of manual vehicles (MVs). Such mixed traffic scenarios with both ICVs and MVs will become increasingly common and the interactions between heterogeneous vehicles will create barriers to operation. Although a single ICV in mixed traffic can identify the interaction intentions of MVs by relying on a single vehicle hardware sensor vehicle computing unit, under the influence of limited computing power and sensing capabilities, as the resource load increases, various errors and risks will be introduced. Additionally, VANET technology under mixed traffic conditions cannot provide complete road resources to match interaction scenarios between ICVs and MVs, and increasing ICV computing demand is increasing the load pressure on VANETs. By adopting the theory of fog computing from the concepts of edge computing, this paper proposes a mixed vehicle fog (MVF) model to leverage the full capabilities of edge nodes in vehicle coupling networks. Based on the reasonable integration and scheduling of ICV resources, the timeliness and security of MV interaction intention calculations can be improved significantly. The proposed model first responds to normal interaction events between ICVs and MVs in a mixed traffic environment based on individual sensing units. It then applies the fault-tolerant node clustering resource scheduling algorithm (FNC-RSA), which dynamically divides ICVs with relevant intention perceptions and calculation requirements for interaction events in local road sections into groups called "collaborative fog groups." Finally, it evaluates ICV node resources and routing costs within the fog groups and schedules resources in a targeted and quantitative manner to realize the sharing of MV interaction information and the collaborative calculation of driving intentions within fog groups. Prescan and MATLAB were used to construct a joint simulation platform for experimentation. The low-energy adaptive clustering hierarchy (LEACH) model was considered for comparison to verify the operation efficiency and robustness of the MVF model. The results demonstrate that the MVF model can subdivide collaborative fog groups based on interactive events, ensure calculation load balancing, improve the calculation and transmission efficiency of ICV directional resources, reduce the average hop number by 55.17%, and reduce the average task completion time by 45.40% compared to the LEACH model. The MVF model also has strong anti-delay interference capabilities and excellent robustness. Therefore, it can play a positive role in breaking the barrier of heterogeneous vehicle interactions in mixed traffic environments, improve the safety of mixed traffic, and create a clean development space for VANETs. |
| |
| Keywords: | traffic engineering MVF model FNC-RSA algorithm intelligent connected vehicle mixed traffic |
|
| 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 |
|
