汽车工程 ›› 2026, Vol. 48 ›› Issue (3): 578-588.doi: 10.19562/j.chinasae.qcgc.2026.03.008

• • 上一篇    

智能网联车辆队列网络拓扑优化及协同控制

高俊1,3,谭小波1,朴昌浩1(),万凯林2   

  1. 1.重庆邮电大学计算机科学与技术学院,重庆 400065
    2.重庆长安汽车软件科技有限公司,重庆 433000
    3.重庆电子科技职业大学智能制造与汽车学院,重庆 401331
  • 收稿日期:2025-04-01 修回日期:2025-05-29 出版日期:2026-03-25 发布日期:2026-03-19
  • 通讯作者: 朴昌浩 E-mail:piaoch@cqupt.edu.cn
  • 基金资助:
    国家重点研发计划项目(2022YFE0101000);重庆市自然科学基金面上项目(2024NSCQ-MSX1981);重庆市教委科学技术研究计划项目(KJON202403124);重庆市教委科学技术研究计划项目(KJQN202503125);重庆市教委科学技术研究计划项目(KJZD-M202503102);2025年教师自主创新“火花”计划(25XIJSCX12)

Network Topology Optimization and Cooperative Control of Intelligent Connected Vehicle Platoons

Jun Gao1,3,Xiaobo Tan1,Changhao Piao1(),Kailin Wan2   

  1. 1.School of Computer Science and Technology,Chongqing University of Posts and Telecommunications,Chongqing 400065
    2.Chongqing Changan Automobile Software Technology Co. ,Ltd. ,Chongqing 433000
    3.School of Intelligent Manufacturing and Automotive Engineering,Chongqing Polytechnic University of Electronic Technology,Chongqing 401331
  • Received:2025-04-01 Revised:2025-05-29 Online:2026-03-25 Published:2026-03-19
  • Contact: Changhao Piao E-mail:piaoch@cqupt.edu.cn

摘要:

针对智能网联汽车通信链路动态变换诱发车辆队列网络拓扑切换,以及传统人工设计拓扑使用场景受限的问题,以实现车辆队列稳定性、舒适性、能耗经济性均衡为目标,提出了一种基于帕累托优化拓扑切换的车辆队列控制方案。首先,基于非支配排序遗传算法,引入网络连通性作为约束条件,以车辆队列3种性能量化指标作为优化目标,离线搜索满足队列性能均衡的帕累托优化网络拓扑;其次,基于分布式非线性模型预测控制器,结合帕累托优化拓扑和马尔科夫链切换机制,设计了基于帕累托优化拓扑切换的车辆队列非线性协同控制器;最后,在3种场景下对车辆队列控制器、优化拓扑筛选方法及切换控制进行了仿真验证,实验结果展示了所提方法的有效性。其中,所提车辆队列控制方法与传统人工设计拓扑切换控制方法相比,其在跟踪稳定性、舒适性以及能耗经济性方面分别提升了30.1%、18.6%和2.2%,表明了本文所提车辆队列协同控制器在动态拓扑切换场景下的可行性和有效性。

关键词: 智能网联汽车, 车辆队列, 帕累托优化, 网络拓扑, 分布式模型预测控制

Abstract:

A Pareto-optimized topology switching control strategy is proposed for connected vehicle platoons to address two key challenges of dynamic changes in communication links that induce network topology switches and the scenario limitation of traditional manually designed topologies. The proposed approach aims to achieve a balanced trade-off among platoon stability, comfort, and energy efficiency. Firstly based on the Non-dominated sorting genetic algorithm, network connectivity is introduced as a constraint, while stability, comfort, and energy efficiency serve as optimization objectives to search offline for Pareto-optimized network topologies that effectively balance these performance metrics. Next, based on a distributed nonlinear model predictive controller, by integrating the Pareto-optimized topology with a Markov chain switching mechanism, a vehicle platoon nonlinear collaborative controller based on Pareto optimization topology switching is designed. Finally, simulation verification is conducted on the vehicle platoon controller, topology selection method, and switching mechanism. The experimental results show the effectiveness of the proposed method. The proposed approach outperforms traditional manually designed topology switching methods, improving tracking stability by 30.1%, comfort by 18.6%, and energy efficiency by 2.2%. These findings demonstrate the feasibility and effectiveness of the proposed cooperative control strategy in dynamic topology switching scenarios.

Key words: intelligent connected vehicles, vehicle platoon, Pareto optimization, network topology, distributed model predictive control