基于改进RRT算法的无人车路径规划仿真研究  

Research on Ground Laser Point Cloud Separation Method Based on Mesh

作　　者：董敏[1] 陈铁桩 杨浩 DONG Min;CHEN Tie-zhuang;YANG Hao(College of Mechanical Engineering,Yanshan University,Qinhuangdao Hebei 066004,China)

机构地区：[1]燕山大学机械工程学院

出　　处：《计算机仿真》2019年第11期96-100,共5页Computer Simulation

基　　金：河北省教育部科学技术研究项目(211024)

摘　　要：路径规划作为无人车核心算法,要规划出合理的避障路径,保证车辆可以在复杂路况上安全行驶。阿克曼转角类型的车辆在行驶过程中受非完整约束,故规划出的路径不仅要符合车辆运动学约束,还要满足算法效率的要求。传统的快速搜索随机树(RRT)算法存在效率低、线路随机等问题,故提出改进的RRT算法。在起点和终点同时生成两棵RRT树,并行计算多条路径,根据评价函数择优选择,并对该路径进行优化和平滑处理。通过仿真和实验的验证,上述方法可以规划出满足车辆控制的要求避障路径,同时具有很高的算法效率,对无人车路径规划算法开发有很好的指导意义。Path planning, as the core algorithm of unmanned vehicle, must plan a reasonable obstacle avoidance path to ensure that vehicles can run safely on complex road conditions. The vehicles of Ackerman angle type are restricted in the course of driving, so the path not only conforms to the kinematic constraints of the vehicle, but also meets the requirements of the efficiency of the algorithm. The traditional fast random-exploring tree(RRT) algorithm has the problems of low efficiency and random lines, so an improved RRT algorithm was proposed. At the same time, two RRT trees were generated at the beginning and end points, and multiple paths were calculated in parallel. According to the evaluation function, the optimal selection was made, and the path was optimized and smoothed. Through the simulation and experimental verification, this method can plan the obstacle avoidance path satisfying the vehicle control and has a high algorithm efficiency. It has a good guiding significance for the development of the unmanned vehicle path planning algorithm.

关 键 词：无人驾驶汽车 算法优化 路径规划 曲线平滑 非完整约束 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]