基于虚拟铰接的V2V无人驾驶车辆编队避障机理与控制研究
批准号:
52072072
项目类别:
面上项目
资助金额:
58.0 万元
负责人:
张宁
依托单位:
学科分类:
运载系统动力学
结题年份:
2024
批准年份:
2020
项目状态:
已结题
项目参与者:
张宁
中文摘要
本项目研究在高速公路行驶工况下,V2V无人驾驶车辆编队的避障机理与控制问题。首次提出虚拟铰接理论与避障安全通道概念。虚拟铰接可使编队在纵向上实现物理硬连接,而在横向上通过虚拟软连接保证每一车辆节点的变道独立可控。避障安全通道是为了在避障中采取异步变道,保证占用最小的横向空间。拟采用从编队整体系统到局部车辆节点、从线性到非线性分析、从分析到综合的方法来解决问题。为此需建立融合车辆动力学、V2V通讯与虚拟铰接理论的编队系统动力学理论模型、虚拟样机模型,采用定性与定量分析方法来研究系统稳定性,揭示其避障机理。确定避障安全通道,为其在避障变道中的路径规划与路径跟踪提供最佳解决方案,并通过分布式控制实现每一车辆节点轮胎力的实时估计与精准控制。借助智能网联技术,采用模型小车试验和道路实车试验来实践避障方案与系统,验证控制方法与算法。研究成果可为完善运载系统动力学与智能化的理论基础做出一定的创新工作。
英文摘要
The research focus of this project is a platoon of driverless vehicles based on Vehicle-to-Vehicle (V2V) communication under the highway working condition. When the platoon encounters obstacles, stability and safety are the first issues to be considered. The physical mechanism of reasonable obstacle avoidance behavior of the platoon is investigated. The control strategy and methods to achieve the behavior are also included in the project. In this project, a new theory of virtual articulation is proposed based on the articulated vehicle system dynamics, and it is applied to the research on the mechanism and control of obstacle avoidance of the V2V platoon. The core idea of virtual articulation theory is to learn from each other's strong points and close the gap. The physical hard connection in the longitudinal direction can be realized by the virtual articulation theory, and it remains stable in the longitudinal direction. In the lateral direction, virtual soft connection is used to ensure that the lane change of each vehicle node is independent and controllable, and the rearward amplification effect caused by lateral hard connection is avoided. The concept of lane change safety corridor (LCSC) is proposed for the first time too. The LCSC is proposed to adopt asynchronous lane change in obstacle avoidance to ensure the minimum lateral space occupation. It is proposed to solve the whole project by using the methods from the whole platoon to the local vehicle nodes, from linear to nonlinear analysis methods, from analysis to synthesis, from theoretical derivation to experimental verification. In the specific project implementation process, the theoretical advantages of virtual articulation concept are taken into full consideration. It is necessary to establish the theoretical model and virtual prototype model in which vehicle dynamics, V2V communication and virtual articulation theory are all integrated. To analyze the platoon stability and string stability in the case of longitudinal, lateral and longitudinal-lateral coupling, the existing analysis methods are still far from enough. Based on the stability analysis methods of articulated vehicles, a new stability analysis method including the criteria is proposed. Then, the obstacle avoidance problem of the V2V platoon is divided into two steps, path planning and path tracking for each vehicle node. The aim of determining the LCSC is to provide the best solution for the two steps. Meanwhile, in the process of obstacle avoidance. the real-time estimation and precise control of the tire forces of each vehicle node is realized through distributed control. In order to practice the obstacle avoidance system and to verify the distributed control strategy, various types of experiments are necessary. With the help of intelligent networking technology, the model car tests and real car tests are planned. Considering that the target working condition is on the highway, the model car tests proposed in the project is divided into three steps. In each step, a model car with a specific chassis is applied with a different driving speed from low to high. These model car tests serve as a feasible and reasonable practical basis for the final real car tests planned in the project. The research results can contribute significant innovations for understanding and improving the theoretical basis of the dynamics and intelligence of the vehicle systems.
本研究以高速公路一字型V2V无人驾驶车辆编队为研究对象,围绕编队避障控制这一核心问题,提出了基于虚拟铰接理论的编队动力学模型及其纵横向协同控制方法,并系统探讨了避障行为的机理与控制策略。研究通过结合队列几何构型与车辆节点动力学的耦合关系,建立了包含轮胎滑移率和侧偏刚度不确定性的编队动力学模型,揭示了低附着力路面等复杂工况对编队稳定性的影响,为后续控制方法的设计提供了理论支撑。.本研究创新性地提出了基于虚拟铰接的编队动力学建模方法,该方法能够有效模拟车辆编队在复杂动态环境中的力学特性,准确描述车辆间的动态相互作用,特别是在避障工况下展现了显著的路径一致性与动态稳定性效果。通过该模型的指导,研究设计了纵横向协同控制方法。纵向控制采用基于李雅普诺夫函数的H∞鲁棒控制方法,能够抑制车辆间的干扰传播并保障编队的弦稳定性;横向控制基于Takagi-Sugeno模糊模型,通过时滞前馈与H∞反馈协同设计,实现了车辆在变道和曲线行驶中的路径跟踪一致性与动态稳定性。.研究进一步提出了针对主动制动与主动转向的编队避障控制方法。设计的纵向鲁棒控制器通过优化虚拟弹簧-阻尼参数,在满足车辆制动力约束的同时,确保纵向稳定性;横向控制通过路径跟踪与模型预测控制器实现车辆编队的同步换道与避障功能。基于MATLAB/Simulink与CarSim仿真实验,以及微缩实验平台的实车测试结果表明,该方法在复杂工况下实现了车辆编队的高效协同与稳定避障。测试结果显示,纵向车距误差小于5%,横向位置误差小于8%,显著提升了车辆编队的安全性与动态性能。.综上所述,本研究通过基于虚拟铰接的动力学建模和纵横向协同控制方法的提出,解决了无人驾驶车辆编队在高速工况下的队列稳定性与避障问题,确保了车辆编队在复杂工况下的协同驾驶的安全性。这些技术突破为智能网联汽车在高速公路等复杂交通环境中的编队行驶和避障能力提供了理论依据与实践指导。
多独立单元协同牵引铰接式重载车辆的瞬态操稳与路径偏移控制研究
-
批准号:52572439
-
项目类别:面上项目
-
资助金额:55.00万元
-
批准年份:2025
-
负责人:张宁
-
依托单位:
汽车–拖车组合系统中车轮摆振与车身摆振的耦合机理研究
-
批准号:51605087
-
项目类别:青年科学基金项目
-
资助金额:20.0万元
-
批准年份:2016
-
负责人:张宁
-
依托单位:
国内基金
海外基金