考虑驾驶风格的网联自动驾驶车辆集聚换道模型

doi:10.13229/j.cnki.jdxbgxb.20250552

摘要/Abstract

摘要：

考虑到未来会出现大量网联自动驾驶车辆（CAV）和人工驾驶车辆（HV）混行的情况，提出了一种考虑驾驶员驾驶风格的集聚换道模型，使CAV实现局部聚集，促进交通流同质化，从而提升交通运行效率。研究驾驶风格在集聚中的影响，定量分析激进、保守以及自动驾驶车辆行驶风格对目标队列集聚（CDA）换道策略、常规集聚（CVA）换道策略、无集聚（NOA）换道策略3种集聚换道模型的影响。基于NGSIM数据库，利用主成分分析法和K-means分析法对车辆驾驶风格进行分类，最后利用Matlab进行仿真模拟，分析3种换道模型下的道路通行能力。研究结果表明，集聚换道模型的应用和激进型驾驶员渗透率的增加有效提高了道路通行能力，所研究的两条道路通行能力最大提升了7%以上，提高保守型驾驶员占比能够提高道路交通流稳定性和安全水平，提高CAV渗透率能够在提高道路通行能力的同时显著提高道路交通流稳定性和安全性。

关键词: 交通运输系统工程, 元胞自动机, 集聚换道模型, 驾驶风格, 自动驾驶, 异质交通流

Abstract:

Considering that a large number of CAV and HV will mix in the future， an agglomeration lane-changing model considering drivers' driving styles was proposed to enable CAV vehicles to achieve local agglomeration， promote traffic flow homogenization， and enhance traffic operation efficiency. The influence of driving styles in agglomeration was investigated， and the effects of aggressive， conservative， and autonomous vehicle driving styles on the three agglomeration lane-changing models CDA， CVA， and NOA were quantitatively analyzed. The agglomeration lane-changing model is based on the NGSIM database， and vehicle driving styles were classified using principal component analysis and K-means analysis， and finally simulated using Matlab to analyze the road capacity under the three lane-changing models. The research results indicate that the application of the lane-changing model and the increase in the penetration rate of aggressive drivers effectively improved road capacity， with the two roads studied showing a maximum increase of more than 7% in capacity. The increase in the proportion of conservative drivers improved road traffic flow stability and safety levels， while the increase in CAV penetration rate significantly improved road traffic flow stability and safety levels while also improving road capacity.

Key words: engineering of communication and transportation, elementary cellular automaton, agglomeration lane-changing model, driving style, autonomous driving, heterogeneous traffic flow

中图分类号:

U491.112

冯天军,郝延铭,李飞燕,高赫遥,刘一贤,刘楠,李金凤. 考虑驾驶风格的网联自动驾驶车辆集聚换道模型[J]. 吉林大学学报(工学版), 2026, 56(7): 1834-1844.

Tian-jun FENG,Yan-ming HAO,Fei-yan LI,He-yao GAO,Yi-xian LIU,Nan LIU,Jin-feng LI. Lane⁃changing model for connected and automated vehicle aggregation considering driving styles[J]. Journal of Jilin University(Engineering and Technology Edition), 2026, 56(7): 1834-1844.

图/表 14

图1

图2

表1

表2

表3

图3

图4

表4

表5

最终聚类中心"

种类	聚类
种类	1	2
$a n$	-1.732 674	0.152 507
$b n$	1.919 921	-0.168 988
$g s a f e$	1.384 303	-0.121 844

表5

表6

表7

车辆模拟参数"

种类	CAV	激进型HV	保守型HV
$v m a x$	20	20	15
$a n$	4	4	2
$b n$	3	3	2
$τ$	0	0.3	0.5
$p s l o w$	0	0.4	0.2
$g s a f e$	1	1.75	3.11

表7

图5

图6

图7

参考文献 22

[1]	Nagel K, Schreckenberg M. A cellular automaton model for freeway traffic[J]. Journal de Physique I, 1992, 2(12): 2221-2229.
[2]	Li X, Wu Q, Jiang R. Cellular automaton model considering the velocity effect of a car on the successive car[J]. Physical Review E, 2001, 64(6): No.066128.
[3]	Chowdhury D, Wolf D E, Schreckenberg M. Particle hopping models for two-lane traffic with two kinds of vehicles: effects of lane-changing rules[J]. Physica A: Statistical Mechanics and its Applications, 1997, 235(3/4): 417-439.
[4]	邱小平, 马丽娜, 周小霞, 杨达. 基于安全距离的手动—自动驾驶混合交通流研究[J]. 交通运输系统工程与信息, 2016, 16(4): 101-108, 124.
	Qiu Xiao-ping, Ma Li-na, Zhou Xiao-xia, Yang Da. The Mixed Traffic Flow of Manual-automated Driving Based on Safety Distance [J]. Journal of Transportation Systems Engineering and Information Technology, 2016, 16(4): 101-108, 124.
[5]	李硕, 夏运达. 改进的考虑前车速度效应的交通流元胞自动机模型[J]. 公路工程, 2019, 44(1): 120-124, 170.
	Li Shuo, Xia Yun-da. Modified cellular automaton models of traffic flow considering the velocity effect of the front vehicle[J]. Highway Engineering, 2019, 44(1): 120-124, 170.
[6]	张治中, 曾俊伟, 钱勇生, 等. 考虑前车安全速度效应的自动驾驶车辆异质交通流特性分析[J]. 交通科技, 2024(1): 96-101.
	Zhang Zhi-zhong, Zeng Jun-wei, Qian Yong-sheng, et al. Heterogeneous traffic flow characterization of autonomous vehicles considering the safe speed effect of the vehicle in front[J]. Transportation Science and Technology, 2024(1): 96-101.
[7]	Wei L, Li W, Liang H, et al. Traffic flow characteristics of speed limited roads based on cellular automata NaSch traffic flow model[C]∥Proceedings of the 4th International Symposium on Computer Science, Digital Economy and Intelligent Systems, Wuhan, China, 2023: 629-638.
[8]	Yan Z, Wang X, Lan B, et al. Improved NS cellular automaton model for simulating traffic flows of two-lane[C]∥IEEE 20th International Conference on Trust, Security and Privacy in Computing and Communications, Shenyang, China, 2021: 1312-1317.
[9]	Kerner B S, Klenov S L, Hermanns G, et al. Effect of driver over-acceleration on traffic breakdown in three-phase cellular automaton traffic flow models[J]. Physica A: Statistical Mechanics and its Applications, 2013, 392(18): 4083-4105.
[10]	武毅. 基于三相交通流理论的元胞自动机模型研究[D]. 长春: 吉林大学交通学院, 2018.
	Wu Yi. Research on a cellular automata traffic flow model for three-phase theory[D] . Changchun: College of Transportation, Jilin University, 2018.
[11]	Feng T, Liu K, Liang C. An improved cellular automata traffic flow model considering driving styles[J]. Sustainability, 2023, 15(2): No.952.
[12]	Liu K, Feng T. Heterogeneous traffic flow cellular automata model mixed with intelligent controlled vehicles[J]. Physica A: Statistical Mechanics and its Applications, 2023, 632: No.129316.
	[[13] Vranken T, Sliwa B, Wietfeld C, et al. Adapting a cellular automata model to describe heterogeneous traffic with human-driven, automated, and communicating automated vehicles[J]. Physica A: Statistical Mechanics and its Applications, 2021, 570: No.125792.
[14]	Shang X-C, Li X-G, Xie D-F, et al. A data-driven two-lane traffic flow model based on cellular automata[J]. Physica A: Statistical Mechanics and its Applications, 2022, 588: No.126531.
[15]	Zheng W, Cheng Y, Ma X, et al. On cellular automata model of traffic flow with Lennard-Jones potential[J]. Nondestructive Testing and Evaluation, 2024, 39(1): 90-101.
[16]	李燕青. 智能网联环境下CAV混行车流集聚控制策略研究[D].镇江:江苏大学汽车与交通工程学院, 2023.
	Li Yan-qing. research on agglomeration control strategy for cav mixed traffic flow in intelligent and connected environment [D]. Zhenjiang: School of Automotive and Transportation Engineering, Jiangsu University, 2023.
[17]	吴德华, 彭锐, 陈荣峰. 异质流网联车的不同换道集聚策略[J]. 西南交通大学学报, 2023, 58(2): 348-356.
	Wu De-hua, Peng Rui, Chen Rong-feng. Hybrid characteristics of heterogeneous traffic flow under different aggregating lane-change strategies in intelligent network[J]. Journal of Southwest Jiaotong University, 2023, 58(2): 348-356.
[18]	蒋阳升, 王思琛, 高宽, 等. 混入智能网联车队的混合交通流元胞自动机模型[J]. 系统仿真学报, 2022, 34(5): No.1025.
	Jiang Yang-sheng, Wang Si-chen, Gao Kuan, et al. Cellular automata model of mixed traffic flow composed of intelligent connected vehicles' platoon[J]. Journal of System Simulation, 2022, 34(5): No.1025.
[19]	商学成. 基于元胞自动机的车辆变道行为建模与交通流特性分析[D] .北京:北京交通大学理学院, 2023.
	Shang Xue-sheng. Lane-changing behavior modeling and traffic flow characteristics analysis based on cellular automaton[D]. Beijing: School of Science, Beijing Jiaotong University, 2023.
[20]	吴德华, 陈荣峰. 智能网联客货混合交通流特性及集聚换道策略[J]. 吉林大学学报: 工学版, 2025, 55 (8): 2588-2596.
	Wu De-hua, Chen Rong-feng. Characteristics of passenger-cargo mixed traffic flow in intelligent network and agglomeration lane-change strategy[J]. Journal of Jilin University(Engineering and Technology Edition), 2025, 55(8): 2588-2596.
[21]	梁军, 李燕青, 王文飒, 等. 智能网联环境下CAV混行车流集聚策略及分析[J]. 华中科技大学学报: 自然科学版, 2024, 52(1): 118-125.
	Liang Jun, Li Yan-qing, Wang Wen-sa, et al. Agglomeration strategy and analysis of CAV mixed traffic flow in intelligent and connected environment[J]. Journal of Huazhong University of Science and Technology(Natural Science Edition), 2024, 52(1): 118-125.
[22]	黄崇钧, 冯天军, 刘福生,等. 基于手机信令数据的公交客流识别[J]. 吉林建筑大学学报, 2025, 42(2): 69-75.
	Huang Chong-jun, Feng Tian-jun, Liu Fu-sheng, et al. Bus passenger flow identification based on mobile signaling data[J]. Journal of Jilin Jianzhu University, 2025, 42(2): 69-75.

相关文章 15

[1]	张玉召,丁欣茹,侯长啸,陈虎林. 交通运输网络韧性研究现状及展望[J]. 吉林大学学报(工学版), 2026, 56(3): 585-602.
[2]	慈玉生,黄轶康. 基于文献计量的交叉口车路协同研究综述[J]. 吉林大学学报(工学版), 2026, 56(2): 313-332.
[3]	赵红星,刘书英,聂江龙,梁瑞艳,何瑞春. 考虑弯道条件的同质与异质交通流稳定性分析[J]. 吉林大学学报(工学版), 2026, 56(2): 393-406.
[4]	张文会,叶梅茹,席聪,宋子文. 混合交通流环境下车辆编队与碳排放特性[J]. 吉林大学学报(工学版), 2026, 56(2): 416-430.
[5]	马健霄,怀硕,赵顗,李铭浩,陈雨欣,赵思雨. 快速路互通交织区超车换道行为控制方法[J]. 吉林大学学报(工学版), 2026, 56(2): 431-442.
[6]	潘义勇,曹天宇,刘宇. 随机交通网络约束最可靠路径乘子交替方向法[J]. 吉林大学学报(工学版), 2026, 56(2): 455-463.
[7]	孙宇,李世武,郭梦竹,金桐彤,宋会军,刘德志,高雯. 多模态数据在驾驶疲劳监测中的有效性分析[J]. 吉林大学学报(工学版), 2026, 56(2): 473-479.
[8]	李坤宸,袁伟,王畅,张会明,穆雨薇. 基于驾驶风险和驾驶能力的人机控制权分配[J]. 吉林大学学报(工学版), 2026, 56(2): 575-584.
[9]	马壮林,毕宇明,周备,邓亚娟,兆雪. 公交换乘优惠政策下居民换乘意向的异质性分析[J]. 吉林大学学报(工学版), 2026, 56(1): 158-169.
[10]	贾洪飞,庄博,罗清玉,刘玲,黄秋阳. 积水条件下城市路网性能恢复决策优化[J]. 吉林大学学报(工学版), 2025, 55(9): 2969-2977.
[11]	王琳虹,刘宇阳,刘子昱,鹿应佳,张宇恒,黄桂树. 基于YOLOv5的轻量化桥梁缺陷识别[J]. 吉林大学学报(工学版), 2025, 55(9): 2958-2968.
[12]	曲昭伟,王铭阳,王喆,宋现敏,张云翔,黄镜尘. 基于自动驾驶模块化车辆主辅功能分配的公交自适应调度方法[J]. 吉林大学学报(工学版), 2025, 55(9): 2946-2957.
[13]	张云翔,宋现敏,谢渝,湛天舒. 基于用户满意度的停车预约服务智能体行为仿真[J]. 吉林大学学报(工学版), 2025, 55(9): 2978-2984.
[14]	穆长儒,徐亮,程国柱. 基于能量合理分配的外包U型钢-混凝土组合护栏防撞性能[J]. 吉林大学学报(工学版), 2025, 55(8): 2669-2680.
[15]	柴树山,周志强,李海涛,徐炅旸. 基于图时空模式学习网络的路网实时交通事件自动检测方法[J]. 吉林大学学报(工学版), 2025, 55(7): 2145-2161.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed

变量	单位	变量	单位
V_mean	m/s	var_dec	m/s²
V_max	m/s	T_mean	m
A_mean	m/s²	var_T	m
D_mean	m/s²	lane_changes	-
var_acc	m/s²

变量	成分
变量	1	2	3
A_mean	0.893 890	0.133 980	0.218 912
D_mean	-0.619 719	-0.652 869	0.335 879
var_acc	0.855 496	0.294 655	0.172 866
var_dec	0.610 886	0.698 697	-0.291 248
var_T	-0.529 064	0.636 036	0.520 895
V_mean	0.725 196	-0.413 668	0.213 898
T_mean	-0.578 801	0.635 256	0.479 241
V_max	0.574 072	-0.372 473	0.583 595
Lane_changes	0.350 848	0.042 036	0.300 216

成分	总计	方差百分比	累积%
1	3.882	43.129	43.129
2	2.139	23.766	66.895
3	1.253	13.921	80.816
4	0.903	10.032	90.848
5	0.372	4.135	94.983
6	0.289	3.211	98.194
7	0.072	0.793	98.989
8	0.048	0.528	99.517
9	0.043	0.483	100.000

代码	实际意义	代码	实际意义
A	平均减速度	F	平均速度
B	时距方差	G	平均加速度
C	最大速度	H	加速度方差
D	平均车头时距	I	减速度方差
E	换道次数

变量	聚类中心
变量	类别1	类别2
A_mean	-0.988 548 407 944 00	0.087 010 356 262 00
D_mean	0.285 272 402 594 00	-0.025 109 299 837 00
var_acc	-0.677 282 431 651 00	0.059 613 284 428 00
var_dec	-0.120 198 726 371 00	0.010 579 803 878 00
var_T	2.858 910 821 477 00	-0.251 636 345 396 00
V_mean	-1.754 638 491 238 00	0.154 440 206 444 00
T_mean	2.885 929 532 673 00	-0.254 014 485 439 00
V_max	-0.901 926 053 876 00	0.079 385 916 648 00
Lane_changes	-0.111 609 498 948 00	0.009 823 678 064 00