【作者】 于康鸿； 张军； 刘元盛； 肖荣春；

【Author】 YU Kanghong;ZHANG Jun;LIU Yuansheng;XIAO Rongchun;Beijing Key Laboratory of Information Service Engineering,Beijing Union University;School of Robotics,Beijing Union University;

【机构】 北京市信息服务工程重点实验室北京联合大学； 北京联合大学机器人学院；

【摘要】 无人驾驶汽车由于其潜在的商业价值与发展前景受到世界各国的广泛关注。自主决策系统是无人驾驶汽车系统中重要的一环,在无人驾驶车辆中充当"大脑"的角色,负责引导无人驾驶车辆无碰撞地从起点出发到达目标点,实质上就是路径规划问题。文中综述了无人驾驶汽车的路径规划方法,以算法的实现方法将无人驾驶汽车路径规划算法分为两大类:传统规划方法和端到端规划方法。文中调研了传统规划方法中代表性的算法与热门的算法,并根据算法的适用范围进行整理与归类,同时调研了端到端方法中热门的方法以及具有代表性的神经网络模型算法。对这些算法的优点与缺点进行了分析和比较,为学习和了解无人驾驶决策技术提供了一个较为全面的参考。更多还原

【Abstract】 Autonomous vehicles have received widespread attention from countries around the world due to their potential commercial value and development prospects.The autonomous decision-making system is an important part of the autonomous vehicle system,playing the role of the "brain" in autonomous vehicles,responsible for guiding autonomous vehicles to start from the starting point to the target point without collision,which is essentially a path planning problem.This paper summarizes the path planning methods for autonomous vehicles,and divides them into two categories based on the implementation methods of algorithms:traditional planning methods and end-to-end planning methods.This paper investigates the classic and popular algorithms in the traditional planning methods,and sorts and classifies them according to the application scope of the algorithms.At the same time,it investigates the popular methods in the end-to-end methods and the representative neural network model algorithms.The advantages and disadvantages of these algorithms are analyzed and compared,which provide a comprehensive reference for learning and understanding autonomous driving decision-making technology.更多还原