CASIA OpenIR  > 毕业生  > 硕士学位论文
Alternative TitleDynamic formation control for multiple Unmanned ground Vehicles based on Cooperative localization
Thesis Advisor赵晓光
Degree Grantor中国科学院大学
Place of Conferral中国科学院自动化研究所
Degree Discipline控制理论与控制工程
Keyword多ugv 协作定位 障碍分类 编队控制 Multi-ugv Cooperative Localization Obstacle Classification Formation Control
Abstract在机器人学的研究中,多机器人研究已经成为当今发展的热门趋势,通过多机器人协作可以解决众多单个机器人无法解决的问题。在多机器人系统中,协作定位技术是决定系统性能的关键,采用多传感器协作定位能够克服使用单个传感器测量精度低、存在累积误差等缺点。野外机器人编队使得机器人能够有效地完成野外追踪、侦查等任务,在协作定位的基础上进行机器人动态编队,并制定野外地形适应策略具有重要的理论意义和实际应用价值。 本论文研究了基于协作定位多UGV动态编队的两个关键技术问题:(1)融合多传感器信息的协作定位;(2)在协作定位的基础上研究多UGV动态编队问题,并且考虑在野外环境中地形适应与相应的编队策略。 首先,为了能够融合多个传感器信息,达到精确定位的目的,本文建立了一个联合滤波模型。该模型由3个子滤波器和一个主滤波器组成,第一个子滤波器利用离散卡尔曼滤波融合光电码盘与陀螺仪信息,推导出惯导系统模型;将第一个子滤波器的结果作为后两个子滤波器的输入,将惯导系统分别与超声传感器信息、GPS信息融合,分别构成了后两个子滤波器。惯导系统与超声传感器融合的滤波器运用两个机器人之间的相对观测信息作为观测输入,运用扩展卡尔曼滤波可以有效的矫正两个机器人的相对位置。惯导系统与GPS信息融合的滤波器将GPS获得的绝对位置信息作为观测输入,通过扩展卡尔曼滤波可以对惯导系统产生的累积误差进行有效矫正。主滤波器按照信息分配的权重融合后两个子滤波器的信息,以达到精确定位的目的。 其次,在使用联合滤波模型对小车定位的基础上,本文提出了野外不确定环境中UGV队形保持的控制方法与队形控制策略。运用基于障碍物分类的通行性方法,通过超声传感器感应障碍物类型,判断小车是否能够通过,然后在此基础上运用人工势场法制定小车的行进策略与队形保持策略。 最后,对协作定位与多UGV编队工作进行了总结与展望。
Other AbstractAmong the study of robotic, multi-robot research has become a hot trend, multi-robot cooperation can solve many problems that single robot can not solve. In multi-robot system, collaboration localization is the key technology to determine the system performance. Multi-sensor collaboration localization can overcome problems like accumulative error, Low measurement precision that single sensor may have. Outdoor robots formation enables the robot to effectively complete the tracking, investigation and other tasks. It has great practical significance to do robot dynamic formation based on collaboration localization and invest wild terrain adaptation strategy. This paper basically has two parts. The first part use multi-sensor to do collaboration localization, and the second part mainly talk about the robot dynamic formation based on collaboration localization and also the wild terrain adaptation strategy. Firstly, in purpose to get precise position of robot using multi-sensor, a joint filter model is constructed which has three sub-filters and a main filter. The first sub-filter use the discrete Kalman filter to fusion the information from photoelectric encoder and gyro, then the Inertial Navigation System model is derived. The output of the first sub-filter is then used as the control input of the other two sub-filters, and the other two sub-sensors fusion the information of Inertial Navigation System with ultrasonic sensors and GPS respectively. The second sub-filter use Relative observation between two robots as observation information, extend Kalman filter can be used to correct the relative position of the two robots. The third sub-sensor use GPS information as observation information, extended Kalman filter can eliminate the accumulated error generated by the Inertial Navigation System. At last, the main filter fusion the information from the last two sub-filter. Secondly, with the basics of robot positioning using collaboration localization, this paper proposes a method for UGV(Unmanned Ground Vehicle) team-formation remain and a controlling strategy in the uncertain outdoor environment. A navigation strategy based on obstacle classification is used. The strategy use an algorithm to decide whether the obstacle detected by laser sensor can be passed through or not, and a formation strategy using the artificial potential field method is proposed based on this. Finally, a Summary and Outlook for cooperative localization multi-UGV formation control is ...
Other Identifier201128014628026
Document Type学位论文
Recommended Citation
GB/T 7714
曾晓腾. 基于协作定位的多UGV动态编队[D]. 中国科学院自动化研究所. 中国科学院大学,2014.
Files in This Item:
File Name/Size DocType Version Access License
CASIA_20112801462802(1621KB) 暂不开放CC BY-NC-SAApplication Full Text
Related Services
Recommend this item
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[曾晓腾]'s Articles
Baidu academic
Similar articles in Baidu academic
[曾晓腾]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[曾晓腾]'s Articles
Terms of Use
No data!
Social Bookmark/Share
All comments (0)
No comment.

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.