双MPC驱动的智能网联交通风险场建模与时空演化

蒋林圆; 丁飞; 樊璇; 杨雪超; 宋爱国; 张登银

doi:10.11999/JEIT260194

双MPC驱动的智能网联交通风险场建模与时空演化

doi: 10.11999/JEIT260194 cstr: 32379.14.JEIT260194

蒋林圆¹,
丁飞^{1, 2, ,},
樊璇¹,
杨雪超¹,
宋爱国³,
张登银¹

1.
南京邮电大学物联网学院南京 210003
2.
南京邮电大学智慧交通学院南京 210003
3.
东南大学空间科学与技术研究院南京 210096

基金项目: 国家自然科学基金(62471241)，江苏省高等学校基础科学(自然科学)研究重大项目(25KJA510003)，江苏省“六大人才高峰”高层次人才资助项目(DZXX-008)，江苏省研究生科研与实践创新计划(SJCX24_0336)

详细信息

作者简介:
蒋林圆：硕士，研究方向为车路协同、智能网联车辆轨迹规划

丁飞：博士，教授，研究方向为群智感知计算、混合信息物理系统

樊璇：硕士，研究方向为自动驾驶路径规划与换道决策

杨雪超：硕士，研究方向为网联车辆风险机理、混合交通流稳定性分析

宋爱国：博士，教授，研究方向为空间探测、智能感知与控制决策

张登银：博士，教授，研究方向为信息处理、模式识别

通讯作者:
丁飞 E-mail:　dingfei@njupt.edu.cn

中图分类号: TP311
计量
- 文章访问数: 15
- HTML全文浏览量: 6
- PDF下载量: 1
- 被引次数: 0
出版历程
- 收稿日期: 2026-02-11
- 修回日期: 2026-04-22
- 录用日期: 2026-04-23
- 网络出版日期: 2026-05-13

Dual-MPC-Driven Modeling and Spatiotemporal Evolution of Intelligent Connected Traffic Risk Fields

1.
School of Internet of Things, Nanjing University of Posts and Telecommunications, Nanjing 210003, China
2.
School of Intelligent Transportation, Nanjing University of Posts and Telecommunications, Najing 210003, China
3.
Institute of Space Science and Technology, Southeast University, Nanjing 210096, China

Funds: The National Natural Science Foundation of China(62471241), The Major Project of the Natural Science Foundation of the Jiangsu Higher Education Institutions of China (25KJA510003), The “Six Talent Peaks” High Level Talent Funding Project of Jiangsu Province (DZXX-008), Jiangsu Graduate Research and Practice Innovation Program (SJCX24_0336)

摘要

摘要: 在智能网联车路云协同系统部署背景下，从路侧视角理清微观车流交互行为与风险场分布的相互作用机理，具有重要理论与应用价值。该文针对多车交互风险的统一建模与动态演化问题，提出一种基于多车动态风险场与双模型预测控制(MPC)的轨迹规划方法。建立面向车路协同的智能网联交通风险场模型，实现驾驶环境风险的连续空间化表征，并引入方向不均匀系数刻画车辆对前向与侧后方风险的感知非对称性。设计基于双MPC的分层决策与运动规划架构，上层基于风险场评估候选驾驶行为，输出最优行为模式；下层根据行为模式激活纵向速度规划MPC或换道轨迹规划MPC，将风险场嵌入代价函数以引导轨迹生成。选用HighD数据集(高速公路)与NGSIM数据集(城市快速路)进行验证，实验结果表明，所提方法能够有效刻画车流风险的时空演化特征，在RMSE、MAE、MAPE等指标上优于现有对比模型，有助于增强路侧系统对局部车流的实时感知与动态风险评估能力。
- 多车交互风险场 /
- 双模型预测控制 /
- 轨迹规划 /
- 车路协同 /
- 风险时空演化
Abstract: Objective With the deployment of intelligent connected vehicle-road-cloud cooperative systems, roadside infrastructure is evolving from traffic-state sensing units into intelligent decision-support platforms for multi-vehicle interaction analysis and dynamic risk inference. In highway and urban freeway scenarios, traffic operation is affected not only by the kinematic responses of individual vehicles but also by lane-changing intentions, car-following competition, and conflict propagation under local interactions. From a roadside perspective, a unified framework is therefore needed to continuously represent traffic risk, reveal its spatiotemporal evolution, and couple risk information with behavior decision-making and trajectory planning. Existing car-following models, such as the Optimal Velocity Model (OVM), Full Velocity Difference (FVD) Model, and Intelligent Driver Model (IDM), can describe speed-spacing evolution. However, these models mainly focus on longitudinal interactions and usually embed risk implicitly in safety-distance or acceleration constraints. They cannot explicitly characterize the coupling between longitudinal following and lateral lane changing, nor can they provide a continuous risk representation suitable for regional traffic assessment. Although Artificial Potential Field (APF) methods and Model Predictive Control (MPC) methods can improve trajectory safety, existing studies still lack a unified mechanism that links risk assessment, behavior decision-making, and motion planning. In addition, discrete behavior choices and continuous control actions are difficult to process efficiently within a single optimization framework. Methods An intelligent connected traffic risk-field model oriented toward vehicle-road cooperation is first established. The model integrates vehicle-interaction risk, lane-marking constraint risk, and road-boundary repulsive risk (Fig. 1). In the vehicle-interaction layer, motion-state-induced risk is formulated by considering the relative speed and relative orientation between the ego vehicle and surrounding vehicles. Distance-induced risk is modeled to reflect attenuation as separation distance increases (Fig. 2(a)–(b)). To represent the stronger influence of forward hazards than lateral and rear hazards, a directional non-uniformity coefficient is used. This coefficient adjusts the angular attenuation of field strength and enables anisotropic spatial risk representation around the vehicle. In the road-constraint layer, lane markings and road boundaries are modeled separately. The total driving risk field is obtained by weighting and combining the lane-marking field, road-boundary field, and multi-vehicle interaction field (Fig. 2(e)–(f)). Based on this representation, a dual-MPC hierarchical decision and motion-planning architecture is designed (Fig. 3). In each control cycle, the upper layer evaluates candidate behavior modes according to the vehicle state, surrounding traffic state, and dynamic risk field, and then outputs a unique behavior mode. The lower layer activates the corresponding control branch. When lane keeping is selected, longitudinal speed-planning MPC is used. When lane changing is selected, lane-change trajectory-planning MPC is activated under road-boundary, lane-marking, and safe-gap constraints. Results and Discussions The proposed framework reconstructs microscopic traffic risk evolution under different datasets and time scales. In the HighD highway scenario, when the slicing interval is 0.4 s, the local evolution of following and lane-changing interactions is captured in detail. This includes the process in which the ego vehicle initially follows a preceding vehicle and then starts changing to the adjacent lane (Fig. 4(a)–(d)). When the interval is increased to 1.0 s, a wider spatiotemporal interaction range becomes visible, and lane-change completion and the subsequent return maneuver are identified more clearly (Fig. 4(e)–(h)). In the NGSIM scenario, a smaller interval provides a finer description of the lane-change disturbance process. By contrast, a larger interval reveals the wider reconstruction of interaction relationships among the original lane, target lane, and surrounding vehicles (Fig. 4(i)–(p)). These results indicate that the proposed roadside-oriented risk field can describe both local interaction details and larger-scale evolution trends, depending on the selected reconstruction interval. Sensitivity experiments further confirm the role of the directional non-uniformity coefficient. As this coefficient increases, forward risk concentration becomes stronger, local peak risk increases, and the coverage of high-risk regions decreases (Table 2). This finding shows that the coefficient effectively regulates anisotropic field distribution. Comparative experiments with IDM, OVM, FVD, and APF show that the proposed method performs better in most representative scenarios and error metrics (Fig. 5, Table 3). In the lateral cut-in scenario, its advantage lies in the early representation of lateral intrusion risk, which enables the behavior decision layer to anticipate conflict and the motion-planning layer to generate continuous evasive actions. In congested scenarios, the superposition of forward congestion risk, lateral neighboring-vehicle influence, and road-boundary constraints allows the dual-MPC controller to evaluate safety and feasibility simultaneously in local space. Conclusions A unified framework for roadside-oriented traffic-risk modeling and behavior-driven trajectory planning is developed. By integrating multi-vehicle interaction risk, lane-marking constraint risk, and road-boundary repulsive risk into a continuously evolving dynamic risk field, the spatial quantification of multi-vehicle interaction risk is realized. The directional non-uniformity coefficient further enables asymmetric risk perception modeling in forward, lateral, and rear directions. On this basis, a dual-MPC hierarchical architecture is constructed to couple behavior decision-making with motion planning, so that lane-keeping and lane-changing behaviors can be adaptively selected and optimized under a unified risk-driven mechanism. Experiments based on HighD and NGSIM datasets show that the proposed method can effectively characterize the spatiotemporal evolution of traffic-risk fields and outperform representative comparison models in most typical scenarios and error metrics.
- multi-vehicle interaction risk field /
- dual model predictive control /
- trajectory planning /
- vehicle-road cooperation /
- spatiotemporal evolution of risk

HTML全文

图 1 智能网联交通风险场建模与时空演化架构

下载: 全尺寸图片幻灯片

图 2 多车交互风险场模型

下载: 全尺寸图片幻灯片

图 3 基于风险场的分层决策与运动规划框架

下载: 全尺寸图片幻灯片

图 4 HighD与NGSIM数据集不同时间间隔下的车流风险场时空演化

下载: 全尺寸图片幻灯片

图 5 不同交通工况下的运动规划轨迹对比

下载: 全尺寸图片幻灯片

表 1 仿真实验参数设置

参数	取值(单位)	参数	取值(单位)
$ {{k}}_{\theta } $	0.7(无量纲)	$ \delta $	1.0(无量纲)
$ {{A}}_{\mathrm{lt}} $	1/exp(14)(无量纲)	$ {{a}}_{\mathrm{max}} $	3.79(m/s²)
$ {{k}}_{\mathrm{L}} $	0.4(无量纲)	$ {{v}}_{\mathrm{desire}} $	100(km/h)
$ {{A}}_{\mathrm{bt}} $	1/exp(14)	$ \mu $	1.25(无量纲)
$ {{k}}_{\mathrm{B}} $	0.4(无量纲)	$ {{T}}_{\mathrm{s}} $	0.1(s)
$ {{N}}_{\mathrm{p}} $	15(无量纲)	$ {{N}}_{\mathrm{e}} $	10(无量纲)

下载: 导出CSV

表 2 HighD与NGSIM数据集下不均匀系数敏感性实验结果

数据集	取值	场景	风险场峰值强度	高风险覆盖率 (%)	RMSE	MAE	MAPE
HighD	0.1	稳定跟驰	1.4758	17.854	1.01010	0.90314	1.6107
		前车减速	1.4758	17.850	1.11610	1.0143	1.2545
		旁车切入	1.4878	16.819	0.40641	0.32538	9.4664
	0.3	稳定跟驰	1.7280	16.819	0.96629	0.86376	1.8554
		前车减速	1.7280	16.825	1.06920	0.97433	1.4686
		旁车切入	1.7749	16.504	0.40641	0.32538	9.4664
	0.5	稳定跟驰	1.9631	15.142	0.92729	0.82856	2.0765
		前车减速	1.9631	15.141	1.02720	0.93867	1.6624
		旁车切入	2.0513	15.175	0.40641	0.32538	9.4664
	0.7	稳定跟驰	2.1722	13.875	0.89420	0.79855	2.2669
		前车减速	2.1722	13.880	0.99142	0.90831	1.8295
		旁车切入	2.3040	13.952	0.40641	0.32538	9.4664
	0.9	稳定跟驰	2.3514	14.127	0.86719	0.77391	2.4244
		前车减速	2.3514	14.113	0.96204	0.88344	1.9679
		旁车切入	2.5253	14.447	0.40641	0.32538	9.4664
NGSIM	0.1	中高速	1.4758	17.833	0.82557	0.71506	0.7326
		低速	1.4750	17.407	0.69200	0.58718	7.0208
		阻塞	1.4753	17.720	4.42710	3.39930	14.5680
	0.3	中高速	1.7280	16.841	0.81314	0.70506	0.8410
		低速	1.7353	16.740	0.69553	0.58957	7.1281
		阻塞	1.7278	16.832	4.47350	3.43370	14.4080
	0.5	中高速	1.9631	15.153	0.80236	0.69649	0.9371
		低速	1.9887	15.141	0.69859	0.59163	7.2317
		阻塞	1.9645	15.151	4.51110	3.46190	14.2280
	0.7	中高速	2.1722	13.890	0.79355	0.68957	1.0174
		低速	2.2281	14.075	0.70111	0.59332	7.3590
		阻塞	2.1763	13.880	4.54360	3.48630	14.0830
	0.9	中高速	2.3513	14.093	0.78664	0.68421	1.0821
		低速	2.4511	14.242	0.70319	0.59471	7.5005
		阻塞	2.3593	14.093	4.57040	3.50640	13.9680

下载: 导出CSV

表 3 不同模型在各场景下的误差指标对比

模型	指标	HighD			NGSIM
模型	指标	稳定跟驰	前车减速	旁车切入	中高速	低速	阻塞
IDM	RMSE	0.8691	1.0035	0.3374	4.1834	0.6517	0.9641
	MAE	0.8690	0.9932	0.3367	4.1802	0.5523	0.8359
	MAPE	0.0172	0.1499	0.2169	2.1758	5.7783	5.3894
OVM	RMSE	1.0365	1.1981	0.3416	5.2766	0.9023	1.2108
	MAE	1.0364	1.1873	0.3408	5.2750	0.7355	1.1384
	MAPE	0.0157	0.1790	1.1543	1.8501	6.7802	5.9028
FVD	RMSE	0.8809	1.0107	0.3391	4.436	0.8880	0.9653
	MAE	0.8808	1.0004	0.3384	4.4354	0.8336	0.8377
	MAPE	0.0171	0.1498	0.4375	1.2184	5.8467	5.3797
APF	RMSE	0.8693	1.0067	0.3383	4.4168	0.6711	0.9812
	MAE	0.8692	0.9963	0.3375	4.4162	0.5733	0.8551
	MAPE	0.0173	0.1492	0.1639	1.2949	5.7647	5.3794
Model (Ours)	RMSE	0.8690	1.0067	0.3373	4.2235	0.6901	0.9667
	MAE	0.8689	0.9965	0.3366	4.2231	0.5877	0.8458
	MAPE	0.0171	0.1481	0.1441	1.0080	5.7198	5.3705

下载: 导出CSV

参考文献(24)

[1]	丁飞, 张楠, 李升波, 等. 智能网联车路云协同系统架构与关键技术研究综述[J]. 自动化学报, 2022, 48(12): 2863–2885. doi: 10.16383/j.aas.c211108. DING Fei, ZHANG Nan, LI Shengbo, et al. A survey of architecture and key technologies of intelligent connected vehicle-road-cloud cooperation system[J]. Acta Automatica Sinica, 2022, 48(12): 2863–2885. doi: 10.16383/j.aas.c211108.
[2]	李立, 李仕琪, 徐志刚, 等. 集成学习框架下的车辆跟驰行为建模[J]. 哈尔滨工业大学学报, 2024, 56(3): 46–55. doi: 10.11918/202205090. LI Li, LI Shiqi, XU Zhigang, et al. Modeling of car-following behavior under an ensemble learning framework[J]. Journal of Harbin Institute of Technology, 2024, 56(3): 46–55. doi: 10.11918/202205090.
[3]	FU Yongchen, HU Wei, SHANG Jian, et al. A fractional-order car-following model with random delay optimized by an enhanced whale optimization algorithm[J]. Nonlinear Dynamics, 2025, 113(16): 21277–21303. doi: 10.1007/s11071-025-11271-4.
[4]	XU Xing, WU Zekun, and ZHAO Yun. An improved longitudinal driving car-following system considering the safe time domain strategy[J]. Sensors, 2024, 24(16): 5202. doi: 10.3390/s24165202.
[5]	QIU Bing and KUANG Hua. A comprehensive review of dynamic modeling methods for extended optimal velocity car-following models in traffic flow[J]. Nonlinear Dynamics, 2025, 113(10): 10861–10879. doi: 10.1007/s11071-024-10660-5.
[6]	ZHANG Jing, GAO Qian, TIAN Junfang, et al. Car-following model based on spatial expectation effect in connected vehicle environment: Modeling, stability analysis and identification[J]. Physica A: Statistical Mechanics and its Applications, 2024, 641: 129747. doi: 10.1016/j.physa.2024.129747.
[7]	VASCONCELOS L and BANDEIRA J M. Calibration of the Intelligent Driver Model (IDM) at the microscopic level[J]. Future Transportation, 2025, 5(2): 57. doi: 10.3390/futuretransp5020057.
[8]	WANG Zhaojie, LU Guangquan, and TAN Haitian. Driving behavior model for multi-vehicle interaction at uncontrolled intersections based on risk field considering drivers’ visual field characteristics[J]. IEEE Transactions on Intelligent Transportation Systems, 2024, 25(11): 15532–15546. doi: 10.1109/tits.2024.3465442.
[9]	GUO Wenfeng, LUO Jing, SONG Xiaolin, et al. A crash-injury-aware driving safety field for real-time risk assessment and its application in autonomous vehicle motion planning[J]. Accident Analysis & Prevention, 2025, 222: 108224. doi: 10.1016/j.aap.2025.108224.
[10]	CHENG Zeyang, ZHU Jinyang, FENG Zhongxiang, et al. Driving safety risk analysis and assessment in a mixed driving environment of connected and non-connected vehicles: A systematic survey[J]. IEEE Transactions on Intelligent Transportation Systems, 2025, 26(5): 5747–5781. doi: 10.1109/tits.2025.3526820.
[11]	WANG Zhuoer, SHI Baohan, ZHANG Jianping, et al. A data-driven spatio-temporal driving risk field mechanism for path planning[J]. Expert Systems with Applications, 2026, 298: 129834. doi: 10.1016/j.eswa.2025.129834.
[12]	WANG Tao, QU Dayi, WANG Kedong, et al. Risk-aware lane change and trajectory planning for connected autonomous vehicles based on a potential field model[J]. World Electric Vehicle Journal, 2024, 15(11): 489. doi: 10.3390/wevj15110489.
[13]	GAO Yan, LI Dazhi, SUI Zhen, et al. Trajectory planning and tracking control of autonomous vehicles based on improved artificial potential field[J]. IEEE Transactions on Vehicular Technology, 2024, 73(9): 12468–12483. doi: 10.1109/tvt.2024.3389054.
[14]	YIN Jiacheng, LI Zongping, CAO Peng, et al. Car-following modeling based on Morse model with consideration of road slope in connected vehicles environment[J]. Physica A: Statistical Mechanics and its Applications, 2023, 622: 128827. doi: 10.1016/j.physa.2023.128827.
[15]	唐伦, 戴军, 成章超, 等. 基于数字孪生的多自动驾驶车辆分布式协同路径规划算法[J]. 电子与信息学报, 2024, 46(6): 2525–2532. doi: 10.11999/JEIT230678. TANG Lun, DAI Jun, CHENG Zhangchao, et al. Distributed collaborative path planning algorithm for multiple autonomous vehicles based on digital twin[J]. Journal of Electronics & Information Technology, 2024, 46(6): 2525–2532. doi: 10.11999/JEIT230678.
[16]	WANG Zhaojie, LU Guangquan, TAN Haitian, et al. A risk-field based motion planning method for multi-vehicle conflict scenario[J]. IEEE Transactions on Vehicular Technology, 2024, 73(1): 310–322. doi: 10.1109/tvt.2023.3308912.
[17]	FANG Xuan and TETTAMANTI T. Change in microscopic traffic simulation practice with respect to the emerging automated driving technology[J]. Periodica Polytechnica Civil Engineering, 2022, 66(1): 86–95. doi: 10.3311/ppci.17411.
[18]	赵川斌, 许伟华, 林博, 等. 融合视觉的多模态通信感知一体化关键技术及原型验证[J]. 电子与信息学报, 2026, 48(2): 487–498. doi: 10.11999/JEIT250685. ZHAO Chuanbin, XU Weihua, LIN Bo, et al. Vision enabled multimodal integrated sensing and communications: Key technologies and prototype validation[J]. Journal of Electronics & Information Technology, 2026, 48(2): 487–498. doi: 10.11999/JEIT250685.
[19]	霍如, 吕科呈, 黄韬. 车联网中路径预测驱动的任务切分与计算资源分配方法[J]. 电子与信息学报, 2025, 47(10): 3658–3669. doi: 10.11999/JEIT250135. HUO Ru, LÜ Kecheng, and HUANG Tao. Task segmentation and computing resource allocation method driven by path prediction in internet of vehicles[J]. Journal of Electronics & Information Technology, 2025, 47(10): 3658–3669. doi: 10.11999/JEIT250135.
[20]	SAMADA S E, PUIG V, NEJJARI F, et al. Safe planning using mixed-integer programming for autonomous vehicles coordination[J]. Robotics and Autonomous Systems, 2025, 193: 105078. doi: 10.1016/j.robot.2025.105078.
[21]	YANG Weishan, CHEN Yuepeng, and SU Yixin. A double-layer model predictive control approach for collision-free lane tracking of on-road autonomous vehicles[J]. Actuators, 2023, 12(4): 169. doi: 10.3390/act12040169.
[22]	WANG Xuefang, JIANG Jingjing, and CHEN Wenhua. High-level decision making in a hierarchical control framework: Integrating HMDP and MPC for autonomous systems[J]. IEEE Transactions on Cybernetics, 2025, 55(4): 1903–1916. doi: 10.1109/tcyb.2025.3535159.
[23]	LIU Ming, ZHANG Yuanjian, LENG Bo, et al. Time-efficient hierarchical vehicle control framework for path tracking in intelligent vehicles[J]. International Journal of Automotive Technology, 2026, 27(2): 595–607. doi: 10.1007/s12239-025-00267-z.
[24]	MA Haozhan, AN Bocheng, LI Linheng, et al. Anisotropy safety potential field model under intelligent and connected vehicle environment and its application in car-following modeling[J]. Journal of Intelligent and Connected Vehicles, 2023, 6(2): 79–90. doi: 10.26599/jicv.2023.9210006.