轮式移动机器人自主定位方法研究
发布时间：2018-11-17 18:09
【摘要】：准确定位是移动机器人实现导航的前提,因此开展移动机器人的自主定位研究具有重要意义。本文对轮式移动机器人在室内环境下的相对定位方法进行了系统地分析研究,主要内容与结论如下：(1)搭建了轮式移动机器人实验平台,实验证明该平台能够满足本文拟定的实验需求。为了分析该实验平台的航迹推测,建立了实验平台的误差分析模型。(2)基于实验平台误差分析模型,分析了轮径与轮距对轮式移动机器人系统误差的影响,并进行了UMBmark校核实验。实验结果表明UMBmark校核算法能有效校地核轮式移动机器人的系统参数,且修正后的系统参数能够明显改善轮式移动机器人的位置估计精度。(3)基于实验平台误差分析模型,分析了轮子打滑对轮式移动机器人随机误差的影响,并利用MEMS陀螺仪与编码器的测量信息对实验平台直线运动时建立了轮子打滑模型,并给出了轮子打滑的判别式以及打滑后移动机器人实际位移的校核式。实验结果表明该方法能准确判别驱动轮是否打滑,同时对驱动轮打.滑校核后能有效提高轮式移动机器人的定位精度。(4)采用g-权重Hough变换与“平面有效区域”集合方法,实现了激光雷达测量与已存地图的匹配,并运用最小二乘法对轮式移动机器人方向与位置误差进行了校核。实验结果表明该方法能够有效补偿位置估计误差,尤其采用的点-点最小二乘法使自主定位更为准确。(5)基于数据融合技术,应用扩展卡尔曼滤波(EKF)融合算法实现了航迹推测与激光雷达定位的融合。实验结果表明该融合算法不仅对方向误差和位置误差具有良好的校核效果,而且能够实现位置的精确跟踪。
[Abstract]:Accurate positioning is the premise of mobile robot navigation, so it is of great significance to carry out autonomous localization of mobile robot. In this paper, the relative positioning method of wheeled mobile robot in indoor environment is systematically analyzed and studied. The main contents and conclusions are as follows: (1) the experimental platform of wheeled mobile robot is built. Experiments show that the platform can meet the experimental requirements of this paper. In order to analyze the trajectory of the experimental platform, the error analysis model of the experimental platform is established. (2) based on the error analysis model of the experimental platform, the influence of wheel diameter and wheel distance on the system error of wheeled mobile robot is analyzed. The UMBmark verification experiment was carried out. The experimental results show that the UMBmark method can effectively calibrate the system parameters of the wheeled mobile robot, and the modified system parameters can obviously improve the accuracy of the position estimation of the wheeled mobile robot. (3) based on the error analysis model of the experimental platform, The influence of wheel skidding on the random error of wheeled mobile robot is analyzed, and the wheel skid model is established by using the measurement information of MEMS gyroscope and encoder for the linear motion of the experimental platform. The discriminant of wheel skidding and the checking of the actual displacement of mobile robot after skid are given. The experimental results show that the proposed method can accurately distinguish whether the driving wheel is sliding and at the same time beat the driving wheel. The accuracy of the wheeled mobile robot can be improved by sliding check. (4) the matching between the lidar measurement and the existing map is realized by using the g- weighted Hough transform and the "plane effective region" set method. The direction and position error of wheeled mobile robot is checked by least square method. The experimental results show that the proposed method can compensate the position estimation error effectively, especially the point-point least square method is used to make the autonomous location more accurate. (5) based on the data fusion technology, The extended Kalman filter (EKF) fusion algorithm is applied to realize the fusion of track estimation and lidar location. The experimental results show that the fusion algorithm not only has a good effect of checking the direction error and the position error, but also can accurately track the position.
【学位授予单位】：北方工业大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TP242
本文编号：2338652
[Abstract]:Accurate positioning is the premise of mobile robot navigation, so it is of great significance to carry out autonomous localization of mobile robot. In this paper, the relative positioning method of wheeled mobile robot in indoor environment is systematically analyzed and studied. The main contents and conclusions are as follows: (1) the experimental platform of wheeled mobile robot is built. Experiments show that the platform can meet the experimental requirements of this paper. In order to analyze the trajectory of the experimental platform, the error analysis model of the experimental platform is established. (2) based on the error analysis model of the experimental platform, the influence of wheel diameter and wheel distance on the system error of wheeled mobile robot is analyzed. The UMBmark verification experiment was carried out. The experimental results show that the UMBmark method can effectively calibrate the system parameters of the wheeled mobile robot, and the modified system parameters can obviously improve the accuracy of the position estimation of the wheeled mobile robot. (3) based on the error analysis model of the experimental platform, The influence of wheel skidding on the random error of wheeled mobile robot is analyzed, and the wheel skid model is established by using the measurement information of MEMS gyroscope and encoder for the linear motion of the experimental platform. The discriminant of wheel skidding and the checking of the actual displacement of mobile robot after skid are given. The experimental results show that the proposed method can accurately distinguish whether the driving wheel is sliding and at the same time beat the driving wheel. The accuracy of the wheeled mobile robot can be improved by sliding check. (4) the matching between the lidar measurement and the existing map is realized by using the g- weighted Hough transform and the "plane effective region" set method. The direction and position error of wheeled mobile robot is checked by least square method. The experimental results show that the proposed method can compensate the position estimation error effectively, especially the point-point least square method is used to make the autonomous location more accurate. (5) based on the data fusion technology, The extended Kalman filter (EKF) fusion algorithm is applied to realize the fusion of track estimation and lidar location. The experimental results show that the fusion algorithm not only has a good effect of checking the direction error and the position error, but also can accurately track the position.
【学位授予单位】：北方工业大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TP242
【参考文献】
相关期刊论文 前1条
1 罗兵;张辉;吴美平;;微机械陀螺同步解调灵敏度分析[J];中国惯性技术学报;2010年02期
相关硕士学位论文 前10条
1 陈二磊;基于嵌入式系统的轮式机器人定位技术的研究与实现[D];西安电子科技大学;2012年
2 张荣昌;轮式移动机器人机构设计与运动控制研究[D];东北大学;2011年
3 李尧;竞速轮式移动机器人系统的开发研究[D];兰州理工大学;2011年
4 洪草根;基于AVR单片机的移动机器人控制系统设计[D];华北电力大学;2011年
5 王兆密;基于MEMS陀螺的惯性测量单元的标定与应用[D];江苏科技大学;2011年
6 张剑光;基于传感器信息权重的移动机器人导航控制[D];辽宁工程技术大学;2011年
7 王慧;智能车的路径识别与控制策略的研究[D];吉林大学;2009年
8 于克;室内自主移动机器人的定位研究[D];燕山大学;2009年
9 张勇;基于多传感器数据融合的清洁机器人定位技术研究[D];哈尔滨工业大学;2008年
10 梁大云;轮式与六杆机构复合移动机器人机械系统研究[D];北京交通大学;2008年
，本文编号：2338652
本文链接：https://www.wllwen.com/kejilunwen/zidonghuakongzhilunwen/2338652.html
