跳到主要內容

臺灣博碩士論文加值系統

(3.235.185.78) 您好!臺灣時間:2021/07/29 23:13
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

: 
twitterline
研究生:傅珮芬
研究生(外文):FU,PEI-FEN
論文名稱:植基於簡單解耦濾波結合最優線性估計器之資訊融合研究
論文名稱(外文):Information Fusion Based on Simple Decoupled Filtering with Optimal Linear Estimator
指導教授:馮力威馮力威引用關係
指導教授(外文):FONG,LI-WEI
學位類別:碩士
校院名稱:育達商業技術學院
系所名稱:資訊管理所
學門:電算機學門
學類:電算機一般學類
論文種類:學術論文
論文出版年:2007
畢業學年度:95
語文別:中文
論文頁數:84
中文關鍵詞:卡爾曼濾波最優線性估計器簡單解耦濾波資訊融合
外文關鍵詞:Kalman FilterOptimal Linear EstimationSimple Decoupled FilteringInformation Fusion
相關次數:
  • 被引用被引用:1
  • 點閱點閱:126
  • 評分評分:
  • 下載下載:23
  • 收藏至我的研究室書目清單書目收藏:0
摘 要
在多感測器網路追蹤系統中利用感測器裡獲得正確的資訊,進而改善追蹤的準確性及即時性是目前的重要研究之一,無論在無人飛行導航或者是雷達追蹤系統中,濾波技術與資訊融合技術皆扮演非常重要的角色。感測器在球面座標系中量測追蹤瞬變目標之距離、方位與俯仰角,需在直角座標系中進行濾波,為了減少計算負荷之實用考量,本研究針對分散式多感測器網路,提出以解耦卡爾曼濾波技術結合α-β-γ濾波器的「簡單解耦濾波(Simple Decoupled Filtering)演算法」,將在視線直角座標系中每一循環計算的濾波增益矩陣,轉換至直角座標系中應用。
在追蹤系統的改進需求方案上,需要精確的狀態估計,感測器資訊融合演算法係由感測器層(Sensor-level)、區域處理器(Local-Processor)與全域處理器(Global-Processor)組合而成,適用在區域慣性直角座標系(Local Initial Cartesian Coordinate System,LICCS)中追蹤單一機動目標。本文研究架構在感測器層使用簡單解耦濾波器進行濾波,在區域處理器中則使用追蹤對追蹤融合(Track to Track Fusion)演算法,將在網路中不同位置的感測器進行組合以獲得個別的追蹤對追蹤融合狀態估計,最後為整合多感測器區域處理器中的各個追蹤對追蹤融合狀態之權重,在全域處理器中使用簡單資訊融合演算法(Simple Information Fusion,SF)、最大概似估計器(Maximum Likelihood Estimator,ML ) 及簡單最大概似估計器(Simplified Maximum Likelihood estimator,SML ) 等三種最優線性估計器(Optimal Linear Estimator)做為資訊融合之方法,並透過電腦模擬驗證方法證明並分析效能。經Monte Carlo電腦模擬結果,驗證「簡單解耦濾波結合最優線性估計器之資訊融合研究」所建議的簡單解耦α-β-γ濾波器,可減少計算量的負荷,提升濾波器的運算速度,且在全域處理器中使用最優線性估計器資訊融合演算法,無論在位置、速度、加速度之估算誤差收歛上均有優異的性能,相較於Sensor-level 平均效能分別提升了約66.86%、52.87%及34.90%;相較於Local-level平均效能分別提升了約42.44%、19.51%及16.77%,證明最優線性估計器資訊融合演算法可明顯改善追蹤精確度。
Abstract
Multi-sensory element network tracking system uses sensory element to obtain the correct information. The improvement in both accuracy and instantaneity are among the present important research. For Unmanned Aerial Navigation or Radar Tracking System, filtering technology and information fusion technology acts as an important role. The distance, azimuth and elevation, measured by the sensory element in the spherical coordinate system, will be filtered in the rectangular coordinate system in order to reduce the computation load for practical consideration. This paper aims the use of filter based on Kalman Decoupled Filtering Integration (or Simple Decoupled Filtering) algorithm. The filter gain formulations are recursively computed in the Line-of-sight Cartesian coordinate system and then transformed for use in the rectangular coordinate system. In this research, a state-vector multi-sensor data fusion approach proposes that consists of sensor-level tracking filters, local- processor and global- processor. This approach is utilized to describe the problem of tracking a maneuvering target in the Local Inertial Cartesian Coordinate System (LICCS).
In the local-processor, the track-to-track fusion algorithm is used to combine the different sensors in network for obtaining individual track-to-track fused state estimate files in the LICCS. In the global-processor, Simple Information Fusion(SF),Maximum Likelihood Estimator(ML),and Simplified Maximum Likelihood Estimator(SML) , three kind of to Optimal Linear Estimator(OLE) to be used as the method of information fusion. Simulation results are presented comparing the performance of the SF、ML and SML algorithm with the Simple Decoupled Filtering algorithm. The ARMSEs of position, velocity and acceleration with the sensor-level filtering algorithm were found to be much larger (about 66.86%, 52.87% and 34.90%) than with the global-processor, respectively. Also, the performance indexes of position, velocity and acceleration with the local-processor were found to be larger (about 42.44%, 19.51% and 16.77%) than with the global-processor, respectively.

Keyword:Kalman Filter, Optimal Linear Estimation, Simple Decoupled Filtering, Information Fusion.
目 錄
指導教授推薦書 ii
論文口試委員審定書 ii
博碩士論文授權書 iii
博碩士論文電子檔案上網授權書 iv
誌 謝 v
摘 要 vi
Abstract viii
圖目錄 xi
表目錄 xiii
表目錄 xiii
符號和縮寫 xiv
第一章 緒論 1
1.1 研究背景與動機 1
1.2 研究目的 5
1.3 論文架構 7
第二章 文獻探討 9
2.1卡爾曼濾波器(Kalman Filter) 9
2.2 簡單解耦α-β-γ濾波器(Simple Decoupled α-β-γ Filter) 14
2.3 資訊融合演算法(Information Fusion) 17
2.3.1追蹤對追蹤資訊融合演算法(Track to Track Fusion) 18
2.3.2 最優線性估計器(Optimal Linear Estimator)資訊融合 19
(1) 簡單資訊融合演算法(Simple Information Fusion) 21
(2) 最大概似估計器(Maximum Likelihood Estimator ) 22
(3) 簡單最大概似估計器(Simplified Maximum Likelihood estimator ) 24
第三章 研究方法 26
3.1 多雷達感測器網路之追蹤定位理論 26
3.1.1 目標動態模型 26
3.1.2感測器量測模式 28
3.2 解耦濾波器之演算法 31
3.3 簡單解耦α-β-γ濾波器之演算法 33
3.4 區域處理器之追蹤對追蹤資訊融合演算法 35
3.5 全域處理器之最優線性估計器演算法 35
第四章 模擬結果與分析 37
4.1 使用(一至五個)感測器之直線模擬測試場景 39
4.2 使用(一至五個)感測器之U型模擬測試場景 46
4.3 使用(一至五個)感測器之圓形模擬測試場景 53
4.2 模擬結果分析 59
第五章 結論與建議 77
參考文獻 79
附錄 .. 86


圖目錄
圖1.1:JDL資訊融合模型 3
圖1.2:分散式多感測器資訊融合架構 6
圖2.1:卡爾曼濾波演算法方程式及濾波流程 13
圖2.2:狀態向量融合法架構圖 18
圖2.3:最優線性估計器架構 20
圖3.1:卡氏直角座標系 28
圖3.2:多感測器α-β-γ Filter簡單資訊融合架構 36
圖4.1(a):直線模擬測試場景,目標、Sensor 1~5之軌跡 40
圖4.1(b):直線模擬測試場景:Local-level與SF、ML、SML之估測位 置均 41
圖4.1(C):直線模擬測試場景:Local-level與SF、ML、SML之估測速度均方根誤差 41
圖4.1(d):直線模擬測試場景:Local-level與SF、ML、SML之估測加速度均方根誤差 42
圖4.2(a):U型模擬測試場景,目標、Sensor 1~5之軌跡 47
圖4.2(b):U型模擬測試場景:Local-level與SF、ML、SML之估測位置均方根誤差 48
圖4.2(c):U型模擬測試場景:Local-level與SF、ML、SML之估測速度均方根誤差 48
圖4.2(d):U型模擬測試場景:Local-level與SF、ML、SML之估測加速度 49
圖4.3(a):圓型模擬測試場景,目標、Sensor 1~5之軌跡 54
圖4.3(b):圓型模擬測試場景:Local-level與SF、ML、SML之估測位置均方根誤差 55
圖4.3(c):圓型模擬測試場景:Local-level與SF、ML、SML之估測速度均方根誤差 55
圖4.3(d):圓型模擬測試場景:Local-level與SF、ML、SML之估測加速度均方根誤差 56
圖4.1(e):直線模擬測試場景:使用二至五個感測器SF、ML和SML之估測位置平均均方根誤差 60
圖4.1(f):直線模擬測試場景:使用二至五個感測器SF、ML和SML之估測速度平均均方根誤差 61
圖4.1(g):直線模擬測試場景:使用二至五個感測器SF、ML和SML之估測加速度平均均方根誤差 61
圖4.2(e):U型模擬測試場景:使用二至五個感測器SF、ML和SML之估測位置平均均方根誤差 65
圖4.2(f):U型模擬測試場景:使用二至五個感測器SF、ML和SML之估測加速度平均均方根誤差 65
圖4.2(g):U型模擬測試場景:使用二至五個感測器SF、ML和SML之估測加速度平均均方根誤差 66
圖4.3(e):圓型模擬測試場景:使用二至五個感測器SF、ML和SML之估測位置平均均方根誤差 70
圖4.3(f):圓型模擬測試場景:使用二至五個感測器SF、ML和SML之估測速度平均均方根誤差 70
圖4.3(g):直線模擬測試場景:使用二至五個感測器SF、ML和SML之估測加速度平均均方根誤差 71
圖4.4(a) 直線模擬測驗場景-以區域處理器為比較基礎之ML、SML及SF估計器於位置、速度與加速度之長條比較圖 77
圖4.4(b) U形模擬測驗場景-以區域處理器為比較基礎之ML、SML及SF估計器於位置、速度與加速度之長條比較圖 78
圖4.4(C)圓形模擬測驗場景-以區域處理器為比較基礎之ML、SML及SF估計器於位置、速度與加速度之長條比較圖 78
附錄 蒙地卡羅電腦模擬流程圖 86

表目錄
表4.1 :分散式多感測器追蹤定位模擬分析. 38
表4.1(a):直線模擬測試場景(一至五個)感測器之模擬結果實驗數據 42
表4.1(b):直線模擬測試場景五個感測器之精度提升比 45
表4.1(a):U型模擬測試場景(一至五個)感測器之模擬結果實驗數據 50
表4.2(b):U型模擬測試場景五個感測器之精度提升比 52
表4.3(a):圓型模擬測試場景(一至五個)感測器之模擬結果實驗數據 58
表4.3(b):圓型模擬測試場景五個感測器之精度提升比 59
表4.1(c):直線模擬測試場景二~五個感測器之整合模擬實驗數據(ARMSE) 62
表4.1(d):直線模擬測試場景感測器提升精度數據(ARMSE) 63
表4.1(e):直線模擬測試場景之位置、速度及加速度效能提升比 64
表4.2(c):U型模擬測試場景二~五個感測器之整合模擬實驗數據(ARMSE) 66
表4.2(d):U型模擬測試場景感測器提升精度數據(ARMSE) 68
表4.2(e):U型模擬測試場景之位置、速度及加速度效能提升比 69
表4.3(c):圓型模擬測試場景二~五個感測器之整合模擬實驗數據(ARMSE) 71
表4.3(d):圓型模擬測試場景感測器提升精度數據(ARMSE) 73
表4.3(e):圓型模擬測試場景之位置、速度及加速度效能提升比 74
表4.4(a):三個場景相較感測器層之效能提升表 75
表4.4(b):三個場景相較區域處理器之效能提升表 75
參考文獻
[1]何友、修建娟、張晶煒、關欣,“雷達數據處理及應用”, 北京電子工業出版社,2000。
[2]韓崇昭、朱洪豔、段戰勝,“多源信息融合”,清華大學出版社。
[3]馮力威、莊謙亮、范承佑、陳中治、傅珮芬,“解耦交互式多模式與加權最小平方估計器應用於資料融合之研究”,育達商業技術學院2007資訊管理及應用國際研討會,pp.147-156, 2007年6月。
[4]馮力威、莊謙亮、傅珮芬“運用多無線通訊地標台解耦資訊融合技術於行動載具定位估測之研究”,樹德科技大學2007資通技術管理與應用會議,2007年6月。
[5]馮力威、陳正雄、樓壁卿、莊謙亮、江冠臻,“簡單解耦雷達追蹤濾波器之設計與分析”,第十二屆國防管理暨實務研討會,2002年6月。
[6]A. Mitch, and J. K. Aggarwal, “Multiple Sensor Integration/Fusion Through Image Processing:A Review”, Optical Engineering, 25(3), pp. 380-386, 1986.
[7]A. R. Riesman, L .W. Nolte “Optimal Detection and Performance of Distribute Sensor Systems.”, IEEE Trans. on Aerospace and Elect. Syst., vo1.AES-23, 24-30, 1987.
[8]A. V. Bal Krishnan, “Kalman Filtering Theory, Optimization Software Inc”, Publications Division, New York, 1984.
[9]B. Edde, “Radar Principles, Technology, Applications”, Prentice Hall, pp. 653-662, 1993.
[10]Y.Bar-Shalom, “On The Track-to-Track Correlation Problems”, IEEE Transactions on Automatic Control, Vol, AC-26, NO.2. April 1981.
[11]Y. Bar-Shalom, X. R. Li., “Multitarget- Multisensor Tracking: Applications and Advances”. Norwood, MA: Artech House, 1990.
[12]Y. Bar-Shalom, “Multitarget-Multisensor Tracking: Applications and Advances Volume II”, MA: Artech House, 1992.
[13]Y. Bar-Shalom, and X. R. Li, “Estimation and tracking principles, techniques”, software. Norwood, MA, Artech House, Chap. 6, pp. 284-287., 1993.
[14]Y. Bar-Shalom, T. E. Fort Mann. “Tracking and Data Association.”, Academic press, 1988.
[15]C. B. Chang, and J. A. Tabaczynski, “Application of State Estimation to Target Tracking” IEEE Transactions on Automatic Control, Vol. AC-29, NO.2, pp. 98-109, 1984.
[16]C. Y. Chong, Spikanta P. Kumar, “Sensor Networks: Evolution, Opportunities, and Challenges”, Proceedings of the IEEE, pp.1247-1256, 2003.
[17]F. Daum, and R. J. Fitzgerald, “ Decoupled Kalman Filter for Phased Array Radar Tracking”, IEEE Transactions on Automatic Control, Vol. AC-28, NO.3, pp. 269-283, 1983.
[18]F. E. White, “Data fusion lexicon. Joint directors of laboratories,” Technical Panel for C3, Data fusion sub-panel, naval ocean systems center, San Diego, CA, USA, 1987.
[19]F. E. White, A. N. Steinberg, C. L. Bowman, “Revision to the JDL DATA Fusion Model. In Sensor Fusion: Architectures, Algorithms, and Applications.” Proceedings of the SPIE. Orlando: Florida, pp.430-441., 1999.
[20]H. Chen, Kirubarajan T., Bar-Shalom Y. “performance limits of track-to-track fusion versus centralized estimation: theory and application.” IEEE Transactions on Aerospace and Electronic Systems, pp.39(2):386-400,2003.
[21]I. Y. Hob Allan, P. K. Varsheny, “Distributed Bayesian Signal Detection”, IEEE Trans. on Info. Theory, pp. 35(5):995-1000, 1989.
[22]J. A. Rocker, and C. D. McGill, “Comparison of two-sensor tracking methods based on state-vector fusion and measurement fusion, ”IEEE Transactions on Aerospace and Electronic Systems, Vol. AES-24, No. 4, pp. 447-449, 1988.
[23]J. B. Pearson, and E. B. Stear, “Kalman Filter Applications in Airborne Radar Tracking,” IEEE Transactions on Aerospace and Electronic Systems, Vol. AES-10, NO.3, pp. 319-329, 1974.
[24]K. C. Chang, R. K. Saha, and Bar-Shalom, Y., “On Optimal Track-to-Track Fusion”, IEEE Transactions on Aerospace and Electronic System, Vol. 33, No.4, October, 1997.
[25]K. V. Ram Chandra, V. S. Srinivasan, “Steady State Results for The x y z Kalman Tracking Filter,” IEEE Transactions on Aerospace and Electronic Systems, Vol. AES-13, pp. 419-423, 1974.
[26]L. W. Fong, P. F. Fu, “Multisensor Track Fusion Algorithms:Covariance Matching Method versus Bayesian-based Estimation.” Proceedings of 2006 CACS Automatic Control Conference, Taipei, Taiwan, Nov. 2006.
[27]L. W. Fong, “Multi-sensor Track-to-Track Fusion Using Simplified Maximum Likelihood Estimator for Maneuvering Target Tracking,” IEEE International Symposium on Industrial Electronics (ISIE 2007), Vigo Spain, pp.36-41., June 2007.
[28]L. W. Fong, P. C. Lou, C. H. Chen and K. Y. Zhen, “Multisensor Fusion via Alpha-Beta-Gamma Filtering with Covariance Matching Method,” 2005 AASRC/CCAS Joint Conference, Kaohsiung Taiwan, Dec. 2005.
[29]L. W. Fong, “Multisensor Fusion via Alpha-Beta-Gamma Filtering with Covariance Matching Method”, Journal of Aeronautics, Astronautics and Aviation, Series A, Vol.38, No.4, pp.241 – 250, 2006.
[30]L. W. Fong, “Distributed Data Fusion Algorithms for Tracking a Maneuvering Target,” The 10th International Conference on Information Fusion, Quebec Canada, July 2007.
[31]M. O. Bed worth, J. Brien. “The Omnibus Model: a new model of data fusion”. IEEE Transactions on Aerospace and Electronic Systems, pp.154 (4):30-36, 2000.
[32]M. Wei , K. P. Schwarz, “Testing a decentralized fitter GPS/INS integration.”, Proc of IEEE PLANS'90, Las Vegas, NV, pp.429-439,1990.
[33]N. A. Carlson, “Federated filter for fault-tolerant integrated navigation system,” Proc. of IEEE PLANS'88, Orlando, pp.110-119. 1988.
[34]P. R. Kalata, “The tracking index: A generalized Parameter for α-β andα-β-γtarget trackers”, IEEE Transactions on Aerospace and Electronic Systems, pp.20 (2), 174-182., 1984.
[35]Q. Gann and C. J. Harris, “Comparison of two measurement fusion methods for Kalman-filter-based multisensor data fusion.” IEEE Transactions on Aerospace and Electronic Systems, Vol. AES-37, NO.1, pp. 273-280, 2001.
[36]R. A. Singer, and A. J. Kanyuck, “Computer control of multiple site track correlation, Automatics”, pp. 455-462, 1971.
[37]R. A. Singer, “Estimating optimum tracking filter performance for manned maneuvering targets,” IEEE Transactions on Aerospace and Electronic Systems, Vol. AES-6,Jul.,pp. 473-483, 1970.
[38]S. J. Julier, J. K. Uhlmann. “A New Method for the Nonlinear Transformation of Means and Covariance in Filters and Estimators.” IEEE Trans. On AC, 45(3), No.3, pp.477-482, 2000:
[39]S. J. Julier, J. K. Uhlmann. “A New Method for the Kalman Filter to Nonlinear systems”. SPIE, Vol.3068, pp.182-193, 1997.
[40]T. L. Song , J. Y. Ahn and C. Park, “suboptimal Filter Design with seudo-measurements for Target Tracking”, IEEE Transactions on Aerospace and Electronic Systems, Vol. 24, NO.1, pp. 28-39, 1988.
[41]V. S. Petrovic , C. S. Xydeas, “Gradient-based Multirresolution Image Fusion”, IEEE Transactions on Image Processing, pp.13(2):287-237,2004.
[42]Y. M. Chen, H. C. Huang, “Multisensor Data Fusion for Maneuvering Target Tracking”, International Journal of Systems Science, Vol. 32, No. 2, pp. 205-214, 2001.
QRCODE
 
 
 
 
 
                                                                                                                                                                                                                                                                                                                                                                                                               
第一頁 上一頁 下一頁 最後一頁 top