智能网联汽车运动规划方法研究综述（双语出版）

分析了近年来智能网联汽车（Intelligent Connected Vehicle，ICV）运动规划方法的研究，根据规划时空尺度和任务目标，将ICV运动规划细分为路径规划、路线规划、动作规划和轨迹规划4级子任务，回顾了各级子任务中智能网联技术的研究和应用现状；探讨了ICV中驾驶人行为特性及其对运动规划结果的影响；从技术背景、研究场景、算法流程和应用理论4个方面，提出ICV运动规划方法研究的未来发展方向。结果表明：由于ICV主要依赖车辆网联信息规划运动路径，而路网中同时存在不同网联等级的ICV，这将增加路径规划问题的求解难度；现有ICV路线规划模型较少考虑周边多车运动状态以及路段车道设置情况，将现有算法与微观交通流模型相结合有助于解决此问题；ICV中人机协同及任务切换领域已出现诸多研究热点，如城市道路上换道与转弯动作规划、ICV引导非网联车辆行驶等问题；借鉴驾驶人行为模式规划ICV运动轨迹已成为研究共识，但是车-车、车-路网联信息在此领域的应用仍然有限；采用反馈-迭代的方法进行ICV运动路线和动作协同规划、运动规划和轨迹跟踪控制有助于获得全局最优的运动规划结果和车辆控制策略；根据具体规划任务特点选择构建ICV运动规划模型的基础理论，有助于发挥各类理论的优势，提升规划算法的灵活性和适用性。

Review of Motion Planning Methods of Intelligent Connected Vehicles（in English）

Recent studies on motion planning methods of intelligent connected vehicle (ICV) are analyzed in this paper. In terms of working space, time, and objective, ICV's motion planning is divided into four subtasks:route planning, path planning, maneuver planning, and trajectory planning. Past research and applications of the techniques of vehicle intelligence and connection in each subtask are reviewed. Behavioral characteristics of the ICV driver and their impact on the outcome of ICV motion planning are discussed. Four aspects of the current trend in ICV motion planning research are discussed:technical background, research scenario, algorithm flow and applied theory. As an ICV mainly depends on vehicle connecting information to plan travelling route, this survey finds that the difficulty of ICV route planning increases when ICV's with different connecting functions coexist in the road network. Dynamics of multiple surrounding vehicles and lane configuration are rarely considered in ICV's path planning. This is likely to be addressed by integrating the existing path planning algorithms with microscopic traffic flow models. The issues of human-machine cooperation and task transfer in ICV have recently become hot topics of research. These issues include lane changing and turning maneuver planning in urban arterial roads, maneuver guidance of ICV for non-connecting vehicle and others. There is academic consensus that the behavior of the driver in an ICV should be considered in trajectory planning. However, there is limited application of vehicle-to-vehicle and vehicle-to-infrastructure connecting information. We propose that applying feedback-iteration to coordinate ICV's path and maneuver planning as well as its motion planning and trajectory tracking control could help in globally optimized motion planning and vehicle control. Furthermore, formulating a model for ICV's motion planning on a theoretical foundation that is appropriate for the specific motion-planning task could not only take advantage of the merits of the theory but also increase flexibility and adaptability of the motion planning algorithms.