CASIA OpenIR  > 毕业生  > 博士学位论文
面向FAST 的多时滞多噪声测量条件下最优状态估计的研究
孙尧
学位类型工学博士
导师梁自泽
2017-04-24
学位授予单位中国科学院研究生院
学位授予地点北京
关键词多时滞 多噪声 最优化 状态估计 Fast工程
其他摘要500 米口径球面射电望远镜 (Five-hundred-meter Aperture Spherical radio
Telescope
,简称 FAST) 是世界上最大单口径和最高观测分辨率的球面射电望
远镜。
FAST 可观测天体的光源信号的波谱范围由馈源舱的跟踪定位精度决定,
而跟踪定位精度很大程度上受到移动馈源舱位置测量精度的影响。移动馈源舱
的测量范围超过
300 米,而其动态测量精度要求达到毫米级,在此种大尺度测
量范围、毫米级动态测量精度的双重要求下,无论
GPS-RTK 还是激光全站仪,
都不能满足
FAST 的要求。本文以 FAST 工程对移动馈源舱位置测量精度的需
求为研究背景,分析常见测量时滞系统状态估计方法无法满足
FAST 精度要求
的原因,详细而深入地研究了多时滞、多噪声测量条件下最优状态估计的若干
问题,主要内容如下:

(1) 围绕双测量源的多时滞多噪声最优估计的相关问题展开研究,使用时间配
准的方法消除测量方程中的显式时滞项,给出了一个能够抑制精度退化的
新型迭代器并用于估计方法中,提出了双测量源的多时滞多噪声状态估计
方法。利用天文观测模式下
FAST 移动馈源舱的实际规划轨迹,模拟馈源
舱沿天文观测轨迹运行时的实时位置测量数据并进行仿真实验,实验表明,
与现有测量时滞系统状态估计方法相比,所提方法提高了估计的精度并且
能够满足
FAST 工程对移动馈源舱位置估计精度的要求。
(2) 将上述状态估计方法拓展至多 (三个及以上) 测量源的情况,围绕广义多时
滞多噪声最优估计问题展开研究,提出了广义多时滞多噪声状态估计方法
的单一引入型和
N 引入型。利用天文观测模式下 FAST 移动馈源舱的实际
规划轨迹,模拟馈源舱沿天文观测轨迹运行时多测量源的实时位置测量数
据并进行仿真实验,实验表明,所提方法能提供高精度的估计结果。

(3) 针对以上多时滞多噪声状态估计的方法,证明了所提方法的最优性和初始
状态无依赖性,为所提方法在工程应用中的可靠性和易用性提供了理论依
据。

; Five-hundred-meter Aperture Spherical radio Telescope (denoted as FAST
for short), is the largest and the most sensitive single dish spherical radio telescope. The effectively observable spectral range of FAST depends on the location
tracking accuracy of the active moving feed cabin. In turn, the location tracking
accuracy of the active moving feed cabin is influenced by its real-time position
measurement accuracy. The maximum measurement range of the active moving
feed cabin is over 300 meters and the required real-time measurement accuracy of
position is RMSE
3 millimeters. However, neither GPS-RTK nor total station,
which are the two selected measurement equipments for the active moving feed
cabin, can meet the demands of FAST. This thesis aims to study MMSE state
estimation for linear discrete-time systems with multi-time-delay and multi-error
measurements, under the research background of the measurement demands of
FAST. The contributions of this thesis are summarized as follows:
(1) A MMSE state estimation approach is designed for linear discrete-time systems with multi-time-delay and multi-error measurements from two measurement channels. This approach uses a time registration method based
innovation re-organization to solve the explicit time-delay problem, and uses
a new iterator to restrain the accuracy degradation phenomenon. Two numerical examples are given to demonstrate the advantage of this approach
and the feasibility of it for FAST.
(2) The above approach is extended to the case of multi (three or more) measurement channels. Therefore, a MMSE state estimation approach is studied for
general linear discrete-time systems with multi-time-delay and multi-error
measurements. One introduction model and
N introduction model of this
approach are designed, respectively. And a numerical example is given to
show the efficiency of these two models.

(3) The above approaches are proved to be optimal state estimation algorithms.
And the optimality is proved to be independent of the assumption on the
initial system state. These two characters provide theoretical foundations
for both the reliability and the usability of the above approaches in actual
applications.

文献类型学位论文
条目标识符http://ir.ia.ac.cn/handle/173211/14727
专题毕业生_博士学位论文
推荐引用方式
GB/T 7714
孙尧. 面向FAST 的多时滞多噪声测量条件下最优状态估计的研究[D]. 北京. 中国科学院研究生院,2017.
条目包含的文件
文件名称/大小 文献类型 版本类型 开放类型 使用许可
thesis.pdf(3967KB)学位论文 暂不开放CC BY-NC-SA请求全文
个性服务
推荐该条目
保存到收藏夹
查看访问统计
导出为Endnote文件
谷歌学术
谷歌学术中相似的文章
[孙尧]的文章
百度学术
百度学术中相似的文章
[孙尧]的文章
必应学术
必应学术中相似的文章
[孙尧]的文章
相关权益政策
暂无数据
收藏/分享
所有评论 (0)
暂无评论
 

除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。