摘要
针对无人驾驶汽车在自动驾驶行进中出现的避障失效、避障规划轨迹不圆滑以及避障算法实时性低等相关动态避障问题,提出一种基于激光雷达的动态避障路径规划方法,通过采用三维激光雷达(Light detection and ranging, LiDAR)实时感知障碍物(人或车)的移动方向、速度以及加速评估障碍物的移动趋势,利用改进的动态规划算法(Dynamic Programming, DP)进行动态障碍物入侵无人车既定路径后的避障路径规划。实验采用百度Apollo无人驾驶开发套件,调用激光雷达并嵌入改进算法以此对基于激光雷达的动态避障路径规划方法验证,实验当有人动态闯入无人车既定路线时,根据障碍物的移动情况,无人车能够及时自主完成车道切换,并避开障碍物按新规划线路行驶,规划的避障路径有效并圆滑、避障规划时间短,符合动态避障的要求。
A dynamic obstacle avoidance path planning method based on Light detection and ranging(LiDAR)is proposed to address the issues of obstacle avoidance failure,non smooth obstacle avoidance planning trajectory,and low real-time performance of obstacle avoidance algorithms in autonomous driving of autonomous vehicles.By using 3D LiDAR to sense the moving direction and speed of obstacles(people or vehicles)in real time and accelerate the e-valuation of the moving trend of obstacles,the improved dynamic programming(DP)algorithm was used to plan the obstacle avoidance path after the dynamic obstacles invade the established path of the unmanned vehicle.The experi-ment used the Baidu Apollo unmanned driving development kit,called on LiDAR and embeded improved algorithms to validate the dynamic obstacle avoidance path planning method based on LiDAR.When someone dynamically enters the established route of the unmanned vehicle,based on the movement of obstacles,the unmanned vehicle can timely and independently complete lane switching,and avoid obstacles to drive along the new planned route.The planned obstacle avoidance path is effective and smooth,and the obstacle avoidance planning time is short,Meet the require-ments of dynamic obstacle avoidance.
作者
马庆禄
张杰
MA Qing-lu;ZHANG Jie(School of Traffic and Transportation,Chongqing Jiaotong University,Chongqing 400074,China)
出处
《计算机仿真》
北大核心
2023年第5期197-201,共5页
Computer Simulation
基金
国家社会科学基金国家应急管理体系建设研究专项(20VYJ023)
重庆市教育委员会教委科学技术研究计划重点项目(KJZD-K202000704)。
关键词
多线激光雷达
无人车
局部轨迹规划
动态障碍物避障
改进的动态规划算法
Multi line LiDAR
Unmanned vehicle
Trajectory planning
Dynamic obstacle avoidance
Improved dy-namic programming algorithm
