中国科学院光电技术研究所机构知识库
Advanced  
IOE OpenIR  > 光电技术研究所博硕士论文  > 学位论文
题名:
光电目标跟踪中的预测滤波技术研究
作者: 熊茂涛
学位类别: 博士
答辩日期: 2009-06-04
授予单位: 中国科学院光电技术研究所
授予地点: 光电技术研究所
导师: 高晓东
关键词: 光电目标跟踪 ; 状态估计 ; 机动目标 ; 可观测性 ; Kalman滤波 ; 粒子滤波
其他题名: Research on Predication Filtering Technology for Opto-electronic Target Tracking
学位专业: 测试计量技术及仪器
中文摘要: 光电目标跟踪中目标状态估计是对跟踪目标选择合适的运动模型和滤波算法,以获得对目标状态的准确估计,并且对目标后续的运动状态做出预测。本论文重点论述了光电目标跟踪测量系统中的预测滤波技术及其实现方法,这对于提高系统的快速捕获能力和跟踪精度具有十分重要的意义。主要研究内容与结果包括以下几个方面。 在总结现有研究成果的基础上,针对光电跟踪测量设备在机动情况、噪声干扰环境下由于目标难以稳定提取而影响跟踪精度与平稳性、实时性的问题,分析了目标跟踪技术的理论基础及其应用现状,在此基础上总结出光电目标跟踪测量中状态估计和预测滤波方面当前需要解决的问题。 针对跟踪目标的复杂性和目标机动运动的随机性,对光电跟踪测量系统中目标的状态估计方法进行了研究,分析了基本的跟踪滤波与预测方法的适应性。通过对光电目标数学模型的分析和研究,建立了光电跟踪测量系统对应的状态模型、观测模型以及目标机动时Kalman滤波方程,讨论了观测数据的获取、噪声分析和模型参数的确定方法,提出了在目标跟踪中利用多次光电目标的观测数据来近似对状态噪声与观测噪声统计特性的确定。此外,在光电跟踪测量系统的数据处理中,通过对机动目标“当前”统计模型的分析,提出了对机动加速度方差的改进计算,进而完成对目标机动强弱的界定分析、跟踪误差的统计评估指标、异常值剔除等问题。 纯方位量测中的可观测性分析是确定最优机动策略的重要问题。针对三维空间作匀加速运动的目标,研究了纯方位量测系统的可观测性问题。利用伪线性化方法和奇异矩阵的一些性质,获得了一种基于可观测性定义和特征值分解法相结合的定量确定纯方位量测系统可观测性的方法,得出系统状态可观测的充要条件。 在对目标状态的预测滤波问题进行深入研究的基础上,把基于自适应滤波原理的滤波调整策略应用于光电跟踪测量设备的预测滤波方法中,在分析和研究目标运动特征和目标模型的基础上,比较了目前比较常用的非线性滤波方法的优缺点及应用情况,包括有EKF、STF、UKF、IMM算法等非线性滤波方法,并且根据各自的应用场合对这些非线性滤波方法进行了改进,提出了一些适用于目标跟踪的高精度预测滤波算法,克服了模型的线性化误差导致的滤波误差较大,滤波发散等问题,提高了数值运算的稳定性,从而使其适合于机动目标运动情况。 利用粒子滤波算法在处理非线性非Gauss目标跟踪问题中的优点,将传统的非线性滤波方法UKF算法和MCMC过程引入到粒子滤波算法中,在重要性密度函数中考虑到了最新的观测值,降低了粒子退化的程度,有效地减轻了样本的贫化现象,做到了滤波精度和数据计算量之间的均衡。 针对光电目标跟踪测量系统的测试需求,利用系统仿真技术建立了机动目标运动轨迹,并应用到光电经纬仪这一类特殊的光电跟踪测量设备中,通过典型的非线性算例仿真实验及实测数据对各种改进的自适应滤波算法进行了验证,分析其运动目标位置、速度、加速度等状态参数的跟踪效果、运行时间以及状态噪声、测量噪声对滤波性能的影响。实验结果的分析表明本论文提出的算法是合理的,能有效地提高收敛速度,改善预测精度,对机动目标和非机动目标均保持了较高精度的跟踪,而且计算量小,做到了滤波精度和数据计算量之间的均衡,是机动目标精确跟踪与定位的优化解决方案,可以为光电跟踪测量设备或者其他类型的跟踪系统的研制和设计提供有益的参考,具有较好的理论和工程实用价值。
英文摘要: The main research to acquire exacting state estimation of opto-electronic targets state includes the selection of the appropriate movement models and filtering algorithms in opto-electronic tracking system and to accurately predicate the state parameters in the following time. The predication filtering technology and the implemental methods for opto-electronic targets tracking are presented in this PhD dissertation, and they are important signification for the targets tracking. The main contents and results are as follows: Aiming at the difficulty of steady tracking and the low accuracy, the low stability and the low real-time property caused by the maneuvering situation and the noise disturbance of the measurement equipments. The essential theory and the application of targets tracking technology are analyzed. The problems needed to be resolved of the field are summed up for the state estimation and the predication filtering. Considering the complexity and randomicity of the opto-electronic maneuvering targets’ movement, the methods of the state estimation are discussed and the adaptability of the essential predication filtering arithmetic is analyzed. The state and the observation models of the opto-electronic targets tracking system and the Kalman filtering equations are established according to the analysis of opto-electronic targets models. The noise source and the acquiring of measurement data and the parameters of models are researched. New method is presented making use of some targets’ observation data in the opto-electronic targets tracking system to approximately acquire the covariance of the state noise and the observation noise. Further, the improved computation of the maneuvering acceleration covariance are presented by analyzing the “current” statistic model, then many problems are solved, such as the confirmation of the targets maneuvering and the statistic evaluation index of the tracking error and the elimination of the exceptional values. The observability analysis is important to confirm the optimal maneuvering strategy. The state equations and measurement equations of targets with constant acceleration in three-dimensional space are founded to study the observability on bearings-only measurement system. Using pseudo-linearity method and character of singularity matrix, the research on the observability is progressed. The observability criterion based on the observability definition and the eigenvalue decomposition solution is acquired. A necessary and sufficient condition of the system deduces from the criterion. The self-adaptive modulating strategies are applied in the predication filtering. Based on the analysis of the targets characteristics and the models, the improved self-adaptive superior precise filtering algorithms are presented by balancing the advantage and disadvantage of the classical nonlinear filtering algorithms. Those filtering algorithms are improved according to respective applicative situations, so they can be applied to the targets maneuvering. Using the advantage of PF arithmetic, the traditional nonlinear filtering algorithms, such as UKF and MCMC method, are introduced to improve PF algorithm. The latest observation data are considered to the important density function. The improved PF algorithm overcomes the degeneration of particles and requires the balancing between the higher tracking precision and the less computation. According to the measured requirement of the opto-electronic tracking system, the movement trajectories of maneuvering targets are established by making use of system emulation technology. Those improved algorithms are applied in the opto-electronic theodolite by using the typical emulation experiments and the real measurement data. The simulation results verify those proposed algorithms and show those self-adaptive improving algorithms can improve the convergence of algorithms and increase the tracking precision and compensates the errors of model linearization and require the proportional balancing between the higher tracking precision and the less computation. Those algorithms are the optimal methods in the targets tracking system as the reference to the tracking system and can be applied to the manufacture and design of practical engineering.
语种: 中文
内容类型: 学位论文
URI标识: http://ir.ioe.ac.cn/handle/181551/345
Appears in Collections:光电技术研究所博硕士论文_学位论文

Files in This Item:
File Name/ File Size Content Type Version Access License
10001_200618015121019熊茂涛_paper.pdf(1528KB)----限制开放View 联系获取全文

Recommended Citation:
熊茂涛. 光电目标跟踪中的预测滤波技术研究[D]. 光电技术研究所. 中国科学院光电技术研究所. 2009.
Service
Recommend this item
Sava as my favorate item
Show this item's statistics
Export Endnote File
Google Scholar
Similar articles in Google Scholar
[熊茂涛]'s Articles
CSDL cross search
Similar articles in CSDL Cross Search
[熊茂涛]‘s Articles
Related Copyright Policies
Null
Social Bookmarking
Add to CiteULike Add to Connotea Add to Del.icio.us Add to Digg Add to Reddit
文件名: 10001_200618015121019熊茂涛_paper.pdf
格式: Adobe PDF
此文件暂不支持浏览
所有评论 (0)
暂无评论
 
评注功能仅针对注册用户开放,请您登录
您对该条目有什么异议,请填写以下表单,管理员会尽快联系您。
内 容:
Email:  *
单位:
验证码:   刷新
您在IR的使用过程中有什么好的想法或者建议可以反馈给我们。
标 题:
 *
内 容:
Email:  *
验证码:   刷新

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

 

 

Valid XHTML 1.0!
Copyright © 2007-2016  中国科学院光电技术研究所 - Feedback
Powered by CSpace