智能化是汽车技术发展的重要方向之一。智能网联车辆编队行驶技术是智能汽车技术研究的一个热点，可以显著提升交通效率、提高道路安全性、改善燃油经济性、降低驾驶劳动强度等。在现有车队控制方案中，跟随车期望路径的生成一般依赖准确绝对定位信息或结构化道路标识，纵向鲁棒控制方法难以保证车间距严格满足安全性不等式约束，横向控制方法缺乏对车辆系统多源不确定性的综合考虑。这些问题使得车队应用场景受限、系统鲁棒性与安全性亟待提升。因此，本文提出了仅依靠车载传感和车车通信的引导式智能网联车队控制方案，该方案可扩展车队应用场景至无绝对定位的非结构化道路，并提升车队系统鲁棒性与安全性。 首先，提出了兼顾避障及前车路径重建的跟随车路径规划方法。基于改进自适应无迹卡尔曼滤波算法，在滑动时间窗内重建前车历史路径点。以历史路径点五次曲线最小二乘拟合为基础，考虑跟随车前方局部区域的障碍物位置，将跟随车路径规划问题建模为凸二次规划问题，得到保证安全性且准确复现前车历史路径的跟随车期望路径。 其次，提出了显式处理车间距安全性不等式约束的车队纵向鲁棒控制方法。为保证车队无碰撞运行，车间距需严格满足安全性不等式约束。针对车辆跟随和车队速度跟随两种模式，分别提出了双射变换技术和车间距误差势函数技术，将车间距不等式约束转化为变换后状态或误差势函数的一致有界性要求，并设计考虑动力学不确定性、执行器饱和及通信时延等因素的鲁棒控制律。该方法可确保车间距满足安全性不等式约束并保证车队稳定性，提升车队纵向控制安全性。 然后，提出了车辆路径跟踪等式约束建模及约束跟随鲁棒控制方法。根据路径跟踪运动学关系构建路径跟踪等式约束，设计基于约束跟随任务空间的不确定性分解方法，得到处于约束跟随任务矩阵零空间和行空间的两部分不确定性。前者不影响约束跟随性能，后者被所设计自适应鲁棒控制律补偿。进一步，提出车辆不确定性模糊集表征方法，并据此设计基于模糊不确定性的路径跟踪鲁棒控制器优化方法，实现了对车辆系统模糊认识下的最优鲁棒控制。 最后，搭建了车队实车试验平台，并在封闭试验场开展了实车试验。试验实现了65km/h 车速下10m 车间距的车辆编队行驶，验证了引导式车队控制方案在提升车队鲁棒性和安全性上的优势。

Intelligence is one of the important directions of the development of vehicular technology. Platooning driving is a hot spot in the research field of intelligent vehicles. It can significantly improve traffic efficiency, road safety and fuel economy, and meanwhile reduce drivers' intensity. However, the existing vehicular platoon control schemes have several problems. The path generation of following vehicles relies on accurate absolute positioning information or structured road signs. The longitudinal robust control methods cannot ensure that the spacing between vehicles strictly satisfies the safety inequality constraints. In addition, the lateral control methods do not comprehensively address the multi-source uncertainties of vehicle system. As a result, the application scenarios of the vehicular platoon are limited, and the robustness and safety of the platoon system need to be improved. To solve these problems, this study proposes a guided intelligent and connected vehicular platoon control scheme that only relies on vehicle on-board sensors and vehicle-to-vehicle communication. This control scheme can expand the application scenarios of intelligent and connected vehicular platoons to unstructured roads without absolute positioning information, and improve the robustness and safety of the platoon system. Firstly, a path planning method that integrates the obstacle avoidance and the front vehicle path reconstruction is proposed for following vehicles. Based on an improved adaptive unscented Kalman filter algorithm, the historical waypoints of the front vehicle are reconstructed in a sliding time window. Under the least square fitting framework of the quintic curve for historical waypoints, considering the obstacle position in the local area in front of the following vehicle, the path planning problem of the following vehicle is modeled as a convex quadratic programming problem. The resulting path ensures the safety of the following vehicle and accurately reproduces the historical path of the front vehicle. Secondly, a robust longitudinal control method that explicitly addresses the safety inequality constraints of inter-vehicle spacing is proposed for the vehicular platoon. In order to realize the collision-free platooning driving, the spacing between vehicles must strictly satisfy the safety inequality constraints. For the vehicle following and platoon speed following driving modes, this study puts forward the bijective transformation technique and the inter-vehicle spacing error potential function technique respectively. The two techniques convert the inequality constraint for the inter-vehicle spacing into the uniform boundedness requirement of the transformed state or the error potential function. On this basis, robust control laws considering the dynamic uncertainty, actuator saturation and communication delay are designed. This method guarantees the satisfaction of the inter-vehicle spacing safety inequality constraints and the stability of the vehicular platoon, and then improves the safety of the platoon longitudinal control. Then, the vehicle path tracking equality constraint modeling and constraint-following robust control methods are proposed. According to the path tracking kinematics, the path tracking equality constraint is formulated, and then an uncertainty decomposition method based on the constraint-following task space is designed, which produces the two parts of uncertainties in the null space and row space of the constraint-following task matrix. The former part does not affect the constraint-following performance, and the latter part is compensated by the designed adaptive robust control law. Furthermore, a fuzzy set modeling method of vehicle uncertainty is proposed, based on which a path tracking robust controller optimization scheme is designed. This optimization scheme realizes the optimal robust control under the fuzzy understanding of the vehicle system. Finally, the real vehicle test platform is built and the experiment verifications are carried out in a closed test field. The experiments realize the platoon driving of vehicles at the speed of 65km/h with 10m spacing. The results also verify the advantages of the guided platoon control scheme in improving the robustness and safety of the platoon system.