本文提出了固定翼無人飛機（UAV）的非線性連續系統模型鑑別方法，並利用黑面琵鷺號無人飛機加以實踐。文中給予無人飛機系統一番詳細的介紹，尤其針對感測器系統多加著墨。這些感測器包括一個用於感測空速的差壓計，兩個安裝在空速管上的攻角（AoA）和側滑角（AoS）測量葉片，以及提供姿態角、角速率和加速度的姿態和航向參考系統（AHRS）。本文的主幹分為三個部份：控制面動態鑑別、感測器數據一致性校正，和非線性力和力矩模型鑑別。控制面動態鑑別是關於控制面真正偏轉角的估算，每一個控制面都由一個伺服馬達驅動，在正常情況下，由於伺服馬達固有的動態特性，以致控制面實際角度有別於指令角度。而伺服馬達的動態則取決於伺服馬達內置的控制系統、伺服馬達和控制面之間的連結，和控制面本身機構。伺服馬達動態鑑別所使用的數據是經由精心設計的實驗收集，並利用觀測器/卡爾曼濾波器鑑別法（OKID）完成鑑別。第二個主要議題是關於感測器一致性校正，本單元討論了感測器的誤差估算，並執行相應的校正使感測器的數據與運動方程式相容。校正過程中主要是使用了RTS平滑器，這個演算法包含了一個擴展型卡爾曼濾波器（EKF）和一個遞歸平滑器。更重要的是，本文運用了溫納濾波器（WF），以避免複雜的雜訊共變異數矩陣估算。此外，本文也引入一個易於實現的大氣數據測量雜訊變異數估算方法。該方法利用全球衛星定位系統（GNSS）所提供的地面速度和上升率以估算大氣數據的雜訊變異數，它包含了數據地域性的想法，並假設附近的數據點之間存在著某種統計上關連。均方根誤差（RMSD）被用於的分析感測器一致性校正結果的好壞。結果顯示，所提出的校正程序很容易實現，而且顯著地提高了無人飛機感測器數據的一致性。最後，本文針對飛機動態的狀態變數和控制輸入對於角加速度和平移加速度的影響進行鑑別，並建立力和力矩數學模型；本文提出的鑑別程序是以拉凡格氏法（LM）為中心，以含有大量式項的模型型態為起始，利用多筆飛測資料進行鑑別，並逐一淘汰不一致的式項，同時找出模型型態和模型參數。鑑別所得到的模型經過不同的飛測資料壓驗證證明模型的可靠性；模型的驗證是以variance accounted for（VAF）為指標。除此之外，筆者也成功地完成了利用飛行數據和非察覺型卡爾曼濾波器（UKF）進行風速估算；風速估算是感測器一致性校正和飛行軌跡重建之間的橋樑，但是這並非本論文的主題，所以會在附錄加以討論。

This dissertation presents a continuous-time nonlinear model identification of a fixed-wing Unmanned Air Vehicle (UAV) called the Spoonbill UAV. An introduction to the UAV system is given to provide a good insight of the onboard avionics, especially on the sensor system. The sensors consist of a differential air pressure transducer for airspeed measurement, two airdata vanes installed on an airdata probe for angle of attack (AoA) and angle of sideslip (AoS) measurement, and an Attitude and Heading Reference System (AHRS) that provides attitude angles, angular rates, and acceleration. The major focus of the dissertation is divided into three parts: control surfaces dynamic model identification, the sensor data compatibility correction, and the nonlinear force and moment models identification. Servo dynamic identification is about estimating the true deflection angle of the control surfaces. The control surfaces are each driven by a servo motor. Under normal circumstances, the control surfaces deflections differ from the commanded/required deflection angles due to the inherent dynamics of the servo motors. The control surfaces dynamics is affected by the built-in control system of the servo motor, the linkage between the servo motors and the control surfaces, and the control surfaces mechanical properties. The control surfaces dynamic model identification was accomplished using data collected in a series of carefully designed experiments and the Observer Kalman Identification (OKID) method. The second major topic is regarding the sensor compatibility correction. This section discusses the estimation of sensor error and performs the corresponding correction such that the sensor data is compatible with the fore equations and kinematic equations. The correction procedure is mainly based on a two pass algorithm called the Rauch-Tung-Striebel (RTS) smoother, which consists of a forward pass Extended Kalman Filter (EKF) and a backward recursion smoother. On top of that, this dissertation proposes the implementation of the Wiener filter prior to the RTS in order to avoid the complicated process noise covariance matrix estimation. Furthermore, an easy to implement airdata measurement noise variance estimation method is introduced. The method estimates the airdata noise variances using the ground speed and ascent rate provided by the Global Navigation Satellite System (GNSS). The method was designed under the assumption that some sort of statistical relation exists between nearby data points. Root mean square deviation (RMSD) is being employed to justify the sensor compatibility. The result shows that the presented procedure is easy to implement, and it improves the UAV sensor data compatibility significantly. Lastly, the nonlinear force and moment models identification is about identifying the mathematical representation of the angular and translational accelerations as functions of the air vehicle dynamic states and control input. The core of proposed identification procedure is the Levenberg-Marquardt (LM) algorithm. The procedure is capable of identifying the model form and the model parameters simultaneously. It was initialized with a general model form with multiple terms. Multiple sets of flight data were used to perform the identification. The most inconsistent term was discarded at each iteration such that the model form took shape slowly. The identified models went through validation using other sets of flight data. The variance accounted for (VAF) was applied to indicate the models' validity. In addition to that, the author has successfully performed the local wind speed estimation using the flight data and Unscented Kalman Filter (UKF). In future, it is possible to combine the work accomplished in the wind estimation and the sensor compatibility correction for flight path reconstruction. However, it is not the main topic of the dissertation and thus discussed in the appendix.

1. Allerton, D., Principles of Flight Simulation, John Wiley & Sons, Wiltshire, UK, pp. 97-154, 2009.
2. Anderson, J.D. Jr., Introduction to Flight, 4th edition, McGraw-Hill, Singapore, pp. 147-157, 312-314, 2000.
3. Balda, M., An algorithm for nonlinear least square, Institute of Thermomechanics, Academy of Sciences of the Czech Republic, 2007.
4. Bamieh, B. and Giarré L., Identification of linear parameter varying models, International Journal of Robust and Nonlinear Control, Vol. 12, pp. 841-853, 2002.
5. Brandon, J.M. and Foster, J.V., Recent dynamic measurements and considerations for aerodynamic modeling of fighter airplane configurations, AIAA Atmospheric Flight Mechanics Conference and Exhibit, Boston, MA, August, 1998.
6. Chaari, A., Elleuch, K., Kharrat, M. and Kamoun, S., Identification of discrete time nonlinear system described by Hammerstein model: Application to a thermal system, ACSE Journal, Vol. 6, No. 2, pp. 63-69, 2006.
7. Chambers, J.R., Modeling Flight: The Role of Dynamic Scaled Free-Flight Models in Support of NASA's Aerospace Programs, National Aeronautics and Space Administration, pp. 139-145, 2010.
8. Chu, Q.P., Mulder, J.A., and van Woerkom, P.Th.L.M., Aircraft flight path reconstruction with nonlinear adaptive filters, In Proceedings of the American Control Conference, Seattle, WA, USA, pp. 1196-1200, June 1995.
9. de Mendonça, C.B., Hemerly, E.M., and Góes, L.C.S., Adaptive stochastic filtering for online aircraft flight path reconstruction, Journal of Aircraft, Vol. 44, pp. 1546-1558, 2007.
10. Eskinat, E. and Johnson, S.H., Use of Hammerstein Models in Identification of Nonlinear Systems, AIChE Journal, Vol. 37, No. 2, pp. 255-268, 1991.
11. Eulrich, B.J. and Weingarten, N.C., Identification and correlation of the F-4E stall/post-stall aerodynamic stability and control characteristics from existing test data, U.S. Air Force Flight Dynamic Laboratory Report, pp. 17-19, 1973.
12. Gavin, H., The Levenberg-Marquardt method for nonlinear least squares curve-fitting problems, Department of Civil and Environmental Engineering, Duke University, April, 2011.
13. Gelb A., Kasper, J.F. Jr., Nash, R.A. Jr., and Price, C.F., and Sutherland, A.A. Jr., Applied Optimal Estimation, The MIT Press, Cambridge, Massachusetts, USA, pp. 156-170, 180-189, 1974.
14. Greblicki, W., Nonparametric identification of Wiener systems by orthogonal series, IEEE Transactions on Automatic Control, Vol. 39, No. 10, pp. 2077-2086, 1994.
15. Hsiao, F.B., Lee, M.T., Cheng, H.C., Lai, Y.C., Chan, W.L., Hsieh, S.Y. and Chen, C.C., SWAN: an autonomous unmanned aerial vehicle project development, The 2nd International Symposium on Innovative Aerial/Space Flyer Systems, Tokyo, Japan, December 2005.
16. Hsiao, F.B., Lee, M.T., Cheng, H.C., Lai, Y.C., Chan, W.L., Hsieh, S.Y. and Chen, C.C., The design and experiment of an autonomous unmanned aerial vehicle: the SWAN project, 2005 AASRC/CCAS Joint Conference, Kaoshiung, Taiwan, December 2005.
17. Hsiao, F.B., Lai, Y.C., Tenn, H.K., Hsieh S.Y., Chen, C.C., Chan, W.L. and Hirst, R., The development of an unmanned aerial vehicle system with surveillance, watch, autonomous flight & navigation capability, 21st International Unmanned Air Vehicle Systems Conference, Bristol, UK, April 2006.
18. Hsiao, F.B., Hsieh, S.Y., Lai, Y.C., Tenn, H.K., Chen, C.C. and Chan, W. L., UAV Velocity Controller Design and Implementation, Proceeding of the 5th Taiwan-Indonesia Workshop on Aeronautical Science, Technology and Industry, Tainan, Taiwan, November 2006.
19. Jang, J.S., Nonlinear control using discrete-time dynamic inversion under input saturation: theory and experiment on the Stanford DragonFly UAVs, Ph.D. Thesis, Stanford University, December 2003.
20. Jategaonkar R.V., Flight Vehicle System Identification: A Time Domain Methodology, American Institute of Aeronautics and Astronautics, Reston, Virginia, USA, pp. 344-350, 2006.
21. Jonker, H.L., Application of the Kalman Filter to flight path reconstruction from flight test data including estimation of instrumental bias error corrections, Report VTH-162, Department of Aerospace Engineering, Delft University of Technology, Delft, The Netherlands, 1976.
22. Juang, J.N., Applied System Identification, Prentice Hall Inc., New Jersey, United States, pp. 175-225, 1994.
23. Julier, S.J. and Uhlmann, J.K., Unscented filtering and nonlinear estimation Proceedings of the IEEE, Vol. 93, No. 3, pp. 401-422, 2004.
24. Klien, V. and Schiess, J.R., Compatibility check of measured air craft responses using kinematic equations and extended Kalman filter, NASA Technical Note TN D-8514, National Aeronautics and Space Administration, Washington, DC, USA, 1977.
25. Klein, V., Morelli, E.A., Aircraft System Identification: Theory and Practice, American Institute of Aeronautics and Astronautics, Reston, Virginia, USA, pp. 27-71, 289-329, 2006.
26. Lai, Y.C., Hsiao, F.B., Development of a low-cost attitude and heading reference system using a three-axis rotating platform, Sensors, Vol.10, pp. 2472-2491, 2010.
27. Lai, Y.C., Hsiao, F.B., Application of fuzzy logic controller and pseudo-attitude to the autonomous flight of an unmanned aerial vehicle, Journal of Chinese Institute of Engineers, Vol. 33, pp. 387-396, 2010.
28. Lee, C.S., The realization of stability augmentation autopilot for unmanned air vehicle, M.Sc. Thesis, National Cheng Kung University, Taiwan, June 2008.
29. Lee, C.S., Hsiao, F.B. and Jan, S.S., Design and implementation of linear-quadratic-Gaussian stability augmentation autopilot for unmanned air vehicle, Aeronaut Journal, Vol. 113, No. 1143, pp. 275-290, 2009.
30. Lee, C.S., Chan, W.L. and Hsiao, F.B., Implementation of system identification on unmanned aerial vehicle via subspace and prediction error method, Journal of Aeronautics, Astronautics, and Aviation, Series A, Vol. 42, No. 2, pp. 87-98, 2010.
31. Lee, C. S., Chan, W. L. , Jan, S. S. and Hsiao, F. B., A linear-quadratic-Gaussian approach for Automatic Flight Control of Fixed-Wing Unmanned Air Vehicles, The Aeronautical Journal, Vol. 115, No. 1163, pp. 29-41, 2011.
32. Lee, C. S., Chan, W. L., Cheng, Y. H. and Hsiao, F. B., Realization of vision-based automatic guidance system on a fixed-wing unmanned aerial vehicle, The Aeronautical Journal. (submitted on May 2011, under review)
33. Ljung, L., System Identification: Theory for the User, Prentice-Hall, New Jersey, USA, pp. 127-140, 1987.
34. Lyshevski, S.E., Identification of nonlinear flight dynamics: theory and practice, IEEE Transaction on Aerospace and Electronic Systems, Vol. 36, No. 2, pp. 383-392, 2000.
35. Marquardt, D., An algorithm for least-squares estimation of nonlinear parameters, Journal of the Society for Industry and Applied Mathematics, Vol. 11, pp. 431-441, 1963.
36. Morelli, E.A., Estimating noise characteristics from flight test data using optimal Fourier smoothing, Jornal of Aircraft, Vol. 32, pp. 689-695, 1995.
37. Murphy, P.C. and Klein, V., Estimation of aircraft unsteady aerodynamic parameters from dynamic wind tunnel testing, AIAA Atmospheric Flight Mechanics Conference and Exhibit, Montreal, Canada, August, 2001.
38. Nelson, R.C., Flight Stability and Automatic Control 2nd Edition, McRaw-Hill, Singapore, pp. 96-127, 1998.
39. Nocedal, J. and Wright, S.J., Numerical Optimization, Springer-Verlag, New York, pp. 259-270, 1999.
40. Owens, D.B., Capone, F.J., Hall, R.M., Brandon, J.M., Cunningham, K. and Chambers, J.R., Free-to-roll analysis of abrupt wing stall on military aircraft at transonic speeds, 41st Aerospace Sciences Meeting and Exhibit, Reno, Nevada, January, 2003.
41. Pottmann, M. and Pearson, R.K., Block-Oriented NARMAX Models with Output Multiplicities, AIChE Journal, Vol. 44, No. 1, pp. 131-140, 1998.
42. Ranganathan, A., The Levenberg-Marquardt Algorithm, Georgia Tech College of Computing, 2004.
43. Rauch, H.E., Tung, F., and Striebel, C.T., Maximum likelihood estimates of linear dynamic systems, AIAA Journal, Vol. 3, pp. 1445-1450, 1965.
44. Rauw, M.O., A SIMULINK environment for flight dynamics and control analysis - application to the DHC-2 'Beaver', Part I, Master Thesis, Delft University of Technology, 1993.
45. Ren, W. and Beard, R.W., Consensus seeking in multi agent systems under dynamically changing interaction topologies, IEEE Transaction on Automatic Control, Vol. 50, No. 5, pp. 655-661, 2005.
46. Stevens, B.L., and Lewis, F.L., Aircraft Control and Simulation, 2nd Edition, John Wiley & Sons, New Jersey, USA, pp. 71-137, 2003.
47. Teo, R., Jang, J.S. and Tomlin, C.J., Automated Multiple UAV Flight - the Stanford DragonFly UAV Program, 43rd IEEE Conference on Decision and Control, Paradise Island, Bahama, December 2004.
48. Todorovic, S. and Nechyba, M.C., A vision system for intelligent mission profiles of micro air vehicles, IEEE Transactions on Vehicular Technology, Vol. 53, No. 6, pp.1713-1725, 2004.
49. Verdult, V. and Verhaegen, M., Bilinear state space systems for nonlinear dynamic modeling, Theory in Biosciences, Vol. 119, No. 1, pp. 1-9, 2000.
50. Verdult, V. and Verhaegen, M., Identification of a Weighted Combination of Multivariable Local Linear State-Space Systems from Input and Output Data, In Proceeding of the 40th IEEE Conference on Decision and Control, Orlando, Florida, USA, December 2001.
51. Verdult, V., Nonlinear system identification: a state-space approach, Ph.D. Thesis, University of Twente, 2002.
52. Wan, E.A. and van der Merwe, R., The unscented Kalman filter for nonlinear estimation, The IEEE 2000 Adaptive Systems for Signal Processing, Communications, and Control Symposium (Lake Louise), pp. 153-58, 2000.
53. Whitmore, S.A., Larson, T.J., and Ehernberger, L.J., Air Data Position-Error Calibration Using State Reconstruction Techniques, NASA Report TM-86029, National Aeronautics and Space Administration, Washington, DC, USA, 1984.
54. Young, B.J., Mobile robots: coordination and control, M.Sc. Thesis, Brigham Young University, April 2000.
55. Crossbow 440 Series Inertial Systems, Crossbow Technology, San Jose, California, USA, pp. 17-27, 2007.
56. U.S. Standard Atmosphere, 1976, National Oceanic and Atmospheric Administration, National Aeronautics and Space Administration, United States Air Force, Washington, DC, USA, pp. 10-12, 1976.