室内移动机器人定位技术研究

发布时间：2019-02-25 18:55

【摘要】：定位是自主移动机器人解决各种复杂任务的关键,对于提高机器人自动化水平具有重要的理论意义和应用价值。机器人定位方法主要分为两大类:相对定位和绝对定位。相对定位,指机器人在初始位置已知的情况下通过传感器来得到自身的位置信息;而绝对定位是机器人在初始位置未知的情况下通过传感器确定自己的位置。本文主要内容包括:首先,本文介绍了机器人定位的基础模块、图像处理基本方法等预备知识,给出了机器人坐标系的建立与变换、机器人的运动学模型和控制算法。其次,本文对相对定位中测程法的机器人模型和定位算法进行研究和分析,搭建了零转弯半径的移动机器人平台,针对机器人的航向信息,加入了电磁罗盘模块,同时引入了卡尔曼滤波算法,使电磁罗盘的角度精确到?1°,为移动机器人提供更加可靠稳定的航向信息,为基于编码器和电磁罗盘的室内移动机器人的定位系统的准确定位奠定了基础,同时零半径转弯使得机器人在运动过程中避障时不会引入额外误差,进一步说明此定位系统的合理性。再次,在绝对定位方面,根据现有资源和室内定位的精度要求,提出了基于计算机视觉的全局定位系统,搭建了156cm×117cm的全局视觉定位平台,通过固定在平台正上方的一个第三视角摄像头对实验平台定期采样,经过一系列图像处理,得到移动机器人顶部两个色标的图像坐标,再根据图像坐标和世界坐标已知的两个参考点,计算得出机器人在实验平台上的世界坐标,得到机器人的位姿信息,从而完成定位。结果表明,视觉全局定位精度达到厘米级别,符合室内移动机器人对定位精度的要求。最后,设计并实现了一种移动机器人上位机——室内移动机器人上位机系统。对于基于编码器和电磁罗盘的定位方法,上位机实时接收机器人的数据,解析出机器人的位置信息,显示在坐标系上。对于视觉全局定位,上位机实时显示摄像头窗口以及处理后的二值图,并定时采样图片进行处理得到机器人的位姿信息,将运动轨迹显示在坐标系上,同时将得到的位姿信息按照无线协议封装显示在数据发送区,并通过性能卓越的无线模块APC220传给机器人,指导其向目标运动。
[Abstract]:Positioning is the key to solve all kinds of complex tasks of autonomous mobile robot, and it has important theoretical significance and application value to improve robot automation level. Robot localization methods are divided into two main categories: relative positioning and absolute positioning. Relative positioning refers to the robot getting its own position information through the sensor when the initial position is known, and absolute positioning is to determine its position through the sensor when the initial position is unknown. The main contents of this paper are as follows: firstly, this paper introduces the basic module of robot positioning, the basic methods of image processing and other preparatory knowledge, gives the establishment and transformation of robot coordinate system, the kinematics model and control algorithm of robot. Secondly, this paper studies and analyzes the robot model and positioning algorithm of the relative location method, builds a mobile robot platform with zero turning radius, and adds the electromagnetic compass module to the direction information of the robot. At the same time, Kalman filter algorithm is introduced to make the angle of electromagnetic compass accurate to-1 掳, which provides more reliable and stable course information for mobile robot. It lays a foundation for accurate positioning of indoor mobile robot based on encoder and electromagnetic compass. At the same time, zero-radius turning makes the robot avoid obstacles in the process of motion without introducing additional errors. The rationality of the positioning system is further explained. Thirdly, in absolute positioning, according to the existing resources and the precision requirements of indoor positioning, a global positioning system based on computer vision is proposed, and a global vision positioning platform based on 156cm 脳 117cm is built. After a series of image processing, the image coordinates of the two color marks on the top of the mobile robot are obtained through a series of image processing, which is fixed directly above the platform by a third angle camera. Then according to the two reference points known to the image coordinates and the world coordinates, the world coordinates of the robot on the experimental platform are calculated, and the position and pose information of the robot is obtained, so that the position and orientation of the robot can be completed. The results show that the global positioning accuracy of vision reaches centimeter level, which accords with the requirements of indoor mobile robot for positioning accuracy. Finally, a kind of mobile robot upper computer-indoor mobile robot upper computer system is designed and implemented. For the positioning method based on encoder and electromagnetic compass, the upper computer receives the data of the robot in real time, analyzes the position information of the robot, and displays it in the coordinate system. For the global positioning of vision, the upper computer displays the camera window and the binary graph after processing in real time, and samples the pictures regularly to get the position and pose information of the robot, and displays the motion track on the coordinate system. At the same time, the position and pose information is displayed in the data transmitting area according to the wireless protocol, and it is transmitted to the robot through the excellent wireless module APC220 to guide it to move toward the target.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP242

【参考文献】