摘要
针对传统自动导引车(Automated guided vehicle,AGV)在室内导航过程中存在定位精度不足,定位成本高,视野单一,以及常规AGV控制灵活性差等问题,本文提出了一种全局视觉的麦克纳姆轮AGV的高精度导航控制系统。全局视觉即在室内高处侧搭载相机并拍摄,通过YOLOv5识别算法和相机标定,获取到AGV以及周围环境物的实时坐标信息。然后将获取到的AGV定位信息与其惯性导航定位进行自适应扩展卡尔曼滤波算法融合,得到更精确的定位信息,最后通过对麦克纳姆轮AGV运动学分析,使用模糊PID以及解耦控制方法,对AGV的运动进行实时纠偏,达到导航控制目的。实验验证,AGV的定位控制精度可达到10 cm以内,定位导航效果得到明显改善。
Aiming at the problems of insufficient positioning accuracy,high positioning cost,single field of view,and poor flexibility of conventional Automated Guided Vehicle(AGV)control in the indoor navigation process,a high-precision navigation control system for Mecanum wheel AGV with global vision is proposed in this paper.Global vision involves mounting a camera at a high location indoors and shoot to form a two-dimensional plane.Firstly,it obtains the real-time coordinate information of the AGV and its surroundings through the recognition algorithm of YOLOv5 and camera calibration.The obtained AGV positioning information is then adaptively fused with its inertial navigation positioning on the AGV using the Extended Kalman Filter algorithm to achieve more accurate positioning.And finally through the analysis of the motion kinematics of the Mecanum wheel AGV,realtime correction of the AGV's motion is achieved using fuzzy PID and decoupling control methods to achieve the navigation control objective.In the final experiment verification,the positioning control accuracy of the AGV can reach within 10 cm,resulting in a noticeable improvement in positioning navigation.
作者
匡兴红
梁定乾
时雨
孔杰
KUANG Xinghong;LIANG Dingqian;SHI Yu;KONG Jie(College of Engineering Science and Technology,Shanghai Ocean University,Shanghai 201306,China;School of Electronic Information and Electrical Engineering,Shanghai Jiao Tong University,Shanghai 200240,China;School of Software,Jiangxi Agricultural University,Nanchang 330045,China)
出处
《机械科学与技术》
北大核心
2025年第12期2069-2077,共9页
Mechanical Science and Technology for Aerospace Engineering
基金
国家重点研发计划(2022YFC3104001)。
