摘要
为提高无人机(Unmanned Aerial Vehicle,UAV)的路径规划能力,设计无人机避障系统。采用深度相机,以Nvidia Xavier NX作为上位机处理与发送指令,采用机器人操作系统(Robot Operating System,ROS)分布式程序设计,实现无人机全自主避障功能。同时采用A^(*)、TEB算法进行路径规划,利用双目相机进行无人机定位,在Ubuntu18.04开发平台中运用Gazebo模拟器建立无人机避障仿真模型,验证项目的可行性与正确性。在所设计的ROS系统控制硬件系统平台进行实际的无人机航线规划、避障导航以及追踪目标试验,结果显示,无人机与障碍物的距离大于0.3 m,所设计的系统可完成避障。
To improve the path planning ability of Unmanned Aerial Vehicles(UAV),this paper designs a UAV obstacle avoidance system.Using a depth camera,the system uses Nvidia Xavier NX as the upper computer to process and send instructions,and adopts a distributed programming design of the Robot Operating System(ROS)to achieve fully autonomous obstacle avoidance tasks for unmanned aerial vehicles.At the same time,A^(*)and TEB algorithms are used for path planning,binocular cameras are used for UAV positioning,and a UAV obstacle avoidance simulation model is established using the Gazebo simulator on the Ubuntu 18.04 development platform to verify the feasibility and correctness of the project.The actual drone route planning,obstacle avoidance navigation,and target tracking experiments were conducted on the designed ROS system control hardware system platform.The results showed that the distance between the drone and the obstacle was greater than 0.3 m,and the designed system could complete obstacle avoidance.
作者
叶霞
谢尚俊
崔浩浩
盛成熠
韩江涛
周诗捷
沈思橙
陈在娥
李胜利
YE Xia;XIE Shangjun;CUI Hahao;SHENG Chengyi;HAN Jiangtao;ZHOU Shijie;SHEN Sicheng;CHEN Zaie;LI Shengli(School of Engineering,Hangzhou Normal University,Hangzhou 310016,China;Qianjiang College,Hangzhou Normal University,Hangzhou 310016,China)
出处
《电视技术》
2023年第5期58-63,共6页
Video Engineering
基金
国家级大学生创新创业训练计划项目(202310346043)
2023年高等教育研究课题(KT2023163)。
