跳到主要內容

臺灣博碩士論文加值系統

(44.211.117.197) 您好!臺灣時間:2024/05/23 12:04
字體大小: 字級放大   字級縮小   預設字形  
回查詢結果 :::

詳目顯示

我願授權國圖
: 
twitterline
研究生:廖翊汶
研究生(外文):Yi-Wen Liao
論文名稱:平行分佈卡門濾波器在GPS/INS統合導引之應用
論文名稱(外文):Parallel Distributed Kalman Filters Applied in GPS/INS Fusion-Based Navigation
指導教授:練光祐
指導教授(外文):Kuang-Yow Lian
學位類別:碩士
校院名稱:中原大學
系所名稱:電機工程研究所
學門:工程學門
學類:電資工程學類
論文種類:學術論文
論文出版年:2001
畢業學年度:89
語文別:英文
論文頁數:138
中文關鍵詞:T-S 模糊模型資料融合卡門濾波器慣性導航系統全球衛星定位系統
外文關鍵詞:T-S fuzzy modelData fusionKalman filterINSGPS
相關次數:
  • 被引用被引用:1
  • 點閱點閱:211
  • 評分評分:
  • 下載下載:0
  • 收藏至我的研究室書目清單書目收藏:1

摘要
本論文提出以T-S模糊模型為基礎的平行分佈卡門濾波器,並利用其做GPS/INS的訊息整合。首先,介紹一些全球衛星定位系統的基本概念。其具有全天候、全球性、及24小時的定位能力之優點,但缺點為量測取樣率低與遮蔽物影響。而慣性導航系統的優點則為量測取樣率高及自主性定位的能力,其缺點為其量測誤差經常時間積分而發散,導致定位精度的降低。基於以上優缺點的考量,因而產生了將GPS/INS訊息做整合的動機,同時亦介紹一些典型的訊息整合架構。接下來的一章,將對卡門濾波器與增廣型卡門濾波器的估測演算法做介紹與討論。並以自走車模型為例,利用增廣型卡門濾波器對其做數值模擬。接著提出本論文的核心部分─T-S模糊平行分佈卡門濾波器。在利用其做系統估測之前,需先將非線性的系統以數個線性的T-S模糊模型表示。第二步,利用一般型卡門濾波器對各個線性的T-S模糊模型分別做狀態估測。第三步,以模糊推論輸出則可獲得原非線性系統之狀態估測。與增廣型卡門濾波器的估測演算法相比,此方法明顯的減低了估測時的運算複雜度。因為一般型卡門濾波器在做估測時,可先得到估測誤差協方差矩陣及卡門增益矩陣,因此在做線上估測時只需做狀態估測即可。但是增廣型卡門濾波器的估測演算法中,上述之矩陣及狀態皆需做線上估測。同樣的,平行分佈卡門濾波器亦對自走車模型做數值模擬。接著提出針對增廣型卡門濾波器的系統雜訊協方差矩陣之調整方法:一、雜訊源以高斯白色雜訊為主;二、雜訊源以系統的不確定項為主。最後,應用平行分佈卡門濾波器於GPS/INS訊息整合。


In this thesis, we propose a Takagi-Sugeno fuzzy model-based parallel distributed Kalman filter for GPS/INS sensor fusion. First some Global Positioning System basic concepts are given. The advantages are all weather, global 24 hour positioning capabilities. The disadvantage is slow update of measurements and the possible loss of signal acquisition do to being obscured. Then important mathematical equations, for example pseusdorange, is given. For inertial navigation systems, the advantages are fast update of sensor measurement and independent positioning without use of external signals. The disadvantage are the integration process may cause divergence due to measurement error. Due to the above facts, we then give the motivations for GPS/INS sensor fusion and some typical fusion structures are shown. In the next part of this thesis, we give a thorough analysis on the Kalman filter estimation algorithm. Both discrete-time standard and extended Kalman filters are compared. Using a mobile robot model as an example, we illustrate the estimation of state for the EKF. Following, we present the main result of this thesis, ``T-S fuzzy parallel distributed Kalman filtering'. To carry out this approach, we first need to model a nonlinear system into T-S fuzzy model local linear representations. In the second step, we use a standard Kalman filter to estimate the state for each linear subsystem. Using the fuzzy inferred output, the overall estimate of state for the original nonlinear system is obtained. This methodology will greatly reduce the computational complexity compared to directly using the EKF due to the fact that standard KF only need to compute the state estimation on-line. In contrast, the EKF computes the state estimation, estimated error covariance, and filter gain online. This parallel distributed Kalman filter is then illustrated in numerical simulations on the mobile robot. A special tuning method for the EKF error covariance matrix is introduced where matrices chosen according to 1) Gaussian white noise; 2) hard bound or soft bound on propagating modeling uncertainty is given. Finally, an experiment of GPS and INS fusion using the parallel distributed Kalman filter with some concepts on positioning and navigation from a practical implementation point of view.


1. Introduction
2. Integration of Global Positioning Systems and Inertial Navigation Systems
3. Kalman Filter Estimation
4. Fuzzy Model-Based Representation and Estimation for Discrete-Time Nonlinear Systems
5. Numerical Simulation Comparisons
6. Positioning/Navigation Concepts and Practical Implementation
7. Conclusions and Future Work
Appendix A. Geodic to Ground Conversion---The Rigorous Proof
Bibliography


1. B. Anderson and J. Moore, Optimal Filtering, Prentice Hall, Englewood Cliffs, New Jersey, 1979.2. J. Borenstein and L. Feng, ``Gyrodometry: a new method for combining data and odometry in mobile robots,'Proc. IEEE ICRA, pp. 423-428, 1996.3. J. Borenstein, H. R. Everett and L. Feng, ``Where am I? Sensors and Methods for mobile robot positioning,' Dept. mechanical engineering and applied mechanics technical report, Univ. Michigan, 1996.4. J. A. Borrie, Stochastic Systems for Engineers: Modeling, Estimation, and Control, Prentice Hall, 1992.5. K. R. Britting, Inertial Navigation Systems Analysis, John Wiley & Sons, Canada, 1971.6. N. A. Carlson, ``Federated filter for fault-tolerant integrated navigation systems,' Proc. IEEE Position Location and Navigation Symposium, pp. 110-119, 1988.7. N. A. Carl, ``Federated square root filter for decentralized parallel processes,' IEEE Trans. Aerospace, Elec. Syst.8. M. Chansarkar and S. Kohli, ``Solution to a multisensor tracking problem with sensor registration errors,' IEEE Trans. Aerospace, Elec., Syst., vl. 35, pp.354-363.9. C. T. Chen, Linear System Theory and Design, New York: Oxford Univ. Press, 1999.10.J. A. Farrell and M. Barth, The Global Positioning System and Inertial Navigation, McGraw-Hill, Ney-York, 1999.

QRCODE
 
 
 
 
 
                                                                                                                                                                                                                                                                                                                                                                                                               
第一頁 上一頁 下一頁 最後一頁 top
無相關期刊