SIA OpenIR  > 水下机器人研究室
基于单目视觉和运动传感器信息融合的水下机器人定位技术研究
Alternative TitleResearch on Positioning Techonology of Underwater Vehicle by Fusing Monocular Vision and Motion Sensors
李强1,2
Department水下机器人技术研究室
Thesis Advisor王晓辉
ClassificationTP242
Keyword水下机器人 数据融合 粒子滤波 相对位置估计
Call NumberTP242/L33/2010
Pages103页
Degree Discipline模式识别与智能系统
Degree Name博士
2010-05-21
Degree Grantor中国科学院沈阳自动化研究所
Place of Conferral沈阳
Abstract论文结合国家863计划项目和中科院创新基金项目开展了基于单目视觉和运动传感器信息融合的水下机器人定位问题研究。提出通过对水下机器人载体的位置和观测目标位置的最优估计来间接获得两者之间相对位置的思路,设计了最优状态估计问题需要的状态方程和测量方程;通过传感器建模获得方程所需参数;通过分析研究现有非线性状态估计方法的不足,提出改进型的PF方法来解决强非线性的最优状态估计问题;提出了简单实用的多速率融合算法将不同采样速率的传感器信息进行融合。仿真试验表明,载体在非零方位速率运动条件下,改进型的多速率PF方法能有效,可靠的估计出水下机器人与目标之间的相对位置,陆上试验表明仅使用惯性测量单元和单目视觉进行数据融合的局限性,水下试验表明DVL有效的弥补了这一局限,可以有效,准确,可靠的估计出载体与目标之间的相对位置。论文研究表明,基于单目视觉和运动传感器的信息融合方法为作业型自治水下机器人提供了一种有效,可行的水下相对位置估计手段,为实现水下机器人的自主作业奠定了基础。
Other AbstractBased on the Chinese National 863 project and the CAS Innovation project, the research in this dissertation is focused on the positioning technology of underwater vehicle by fusing monocular vision and motion sensors. The basic idea of obtaining the relative position between AUV(Autonomous Underwater Vehicle) and target is estimating the position state of AUV and target simultaneously. In the dissertation the state equations and measurement equations which are necessary for optimal state estimation are designed. Some important parameters in these equations are obtained by modeling for sensors. After analyzing the drawback of conventional nonlinear state estimation, an improved PF algorithm is proposed for this special strong nonlinear estimation problem, and a simple and practical multirate fusion algorithm is proposed to fuse the measurement value in different frequency. It can be seen from the simulation that when the vehicle moved with non zero-bearing rate, the relative position can be obtained effectively and reliably by improved multirate PF method. The experiment done on-land showed the limitation of method by fusing monocular vision and IMU (Inertial Measurement Value). The experiment done underwater showed that the DVL(Doppler Velocity Log) can make up for the deficiency and the relative position can be obtained accurately, effectively and reliably by fusing the information from  monocular vision, IMU and DVL. The research in the dissertation showed that fusion method from motion sensors and monocular is an effective and feasible position method for IAUV (Intervention Autonomous Underwater Vehicle) and it lays the foundation for the autonomous intervention of underwater vehicle.
Language中文
Contribution Rank1
Document Type学位论文
Identifierhttp://ir.sia.cn/handle/173321/9271
Collection水下机器人研究室
Affiliation1.中国科学院沈阳自动化研究所
2.中国科学院研究生院
Recommended Citation
GB/T 7714
李强. 基于单目视觉和运动传感器信息融合的水下机器人定位技术研究[D]. 沈阳. 中国科学院沈阳自动化研究所,2010.
Files in This Item:
File Name/Size DocType Version Access License
基于单目视觉和运动传感器信息融合的水下机(1978KB) 开放获取LicenseApplication Full Text
Related Services
Recommend this item
Bookmark
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[李强]'s Articles
Baidu academic
Similar articles in Baidu academic
[李强]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[李强]'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.