带双目视觉的全向移动物流机器人设计与实现  

Design and Implementation of Omnidirectional Mobile Logistics Robot with Binocular Vision

作　　者：戴月[1] DAI Yue(Training Center,Chuzhou Vocational And Technical College,Chuzhou 239000,China)

机构地区：[1]滁州职业技术学院实训中心,安徽滁州239000

出　　处：《太原学院学报（自然科学版）》2020年第3期35-41,共7页Journal of TaiYuan University:Natural Science Edition

基　　金：2019年校级自然科学重点工程项目(YJZ-2019-05);2019年校级自然科学重点工程项目(YJZ-2019-15)。

摘　　要：随着中国物流工业的发展,仓库机器人的需求也在增加,机器人库存管理逐渐向自动化、智能化的方向发展。在机器人仓库管理系统中,机器人不仅能节省劳动力,还能减少人力资源浪费,提升仓库储存效率。针对物流仓储机器人控制问题,设计一种带有双目视觉传感器的全方位移动机器人,实现产品的仓储物流自动化。分析了国内外相关领域的研究现状,从控制器、导航定位技术、传感器信息融合技术三方面阐述了物流移动机器人的关键技术;分析了全向移动机器人的轮系特点并构建运动学模型;针对全向移动机器人的导航定位问题设计了基于双目视觉的导航系统,给出了摄像机标定的方法,以及双目信息提取和特征匹配融合算法;通过视觉标定实验和视觉导航实验验证了所提出的标定与导航方法的有效性,并拓展导航方法覆盖的环境,使得该方法具有更广泛的实用范围。In recent years,with the development of China's logistics industry,the demand for robotic warehouse robots has also increased,and the robot inventory management system has gradually become more automated and intelligent.In the robotic warehouse management system,robots not only save labor,but also reduce the damage caused by abuse,and also improve storage efficiency.Aiming at the robot control problem in the logistics industry,this paper designs an all-round mobile robot with binocular vision sensor to complete the warehousing logistics automation of the product.The research status of related fields at home and abroad is analyzed,and the key technologies of logistics mobile robots are described in three parts:controller,navigation and positioning technology and sensor information fusion technology.The characteristics of omnidirectional mobile robots are analyzed.The kinematics model is constructed.A navigation system based on binocular vision is designed for the navigation and positioning problem of omnidirectional mobile robots.The method of camera calibration and binocular information extraction and feature matching for navigation information processing are given.Finally,the effectiveness of the proposed calibration and navigation methods is verified by visual calibration experiments and visual navigation experiments.The navigation method covers indoor and an outdoor environment,which makes the method have a wider range of practical applications.

关 键 词：移动机器人 双目视觉 特征匹配 麦克纳姆轮 

分 类 号：TP242.3[自动化与计算机技术—检测技术与自动化装置]