本研究主旨在建立一動態飛行參數驗證程式。文中採用自相關類神經網路(Auto-Associative Neural Networks, AANN)作為程式發展的基礎，再加入迴傳式類神經網路(Recurrent Neural Networks, RNN)的觀念，發展出一套可驗證飛行參數的迴傳式自相關類神經網路(Recurrent AANN, R-AANN)。R-AANN可模擬飛行參數之間的動態關係，掌握各項參數隨時間的變化歷程及相互間關聯性，並透過此R-AANN以判斷各參數是否相容(Compatible)。為測試本網路的推廣能力，本文先藉由簡單的物理模型與非線性離散化模型來測試程式之可行性後，再使用於動態飛行參數上。為了方便後續的驗證，在此所使用之飛行參數是由飛行力學六自由度運動方程式(Six-Degree-of-Freedom Equations of Motion)所產生。結果顯示，本方法使用於動態飛行之縱向參數上有相當不錯之學習能力與相容性鑑定效果。

The objective of the present research is to develop a software system that can perform dynamic flight data validation using artificial neural networks. Auto-associative neural network (AANN) was adopted as the groundwork for method development. In order to strengthen the generalization ability of this AANN when coping with various dynamic situations that are not included in the training scenarios, recurrent neural network (RNN) was incorporated into this data validation AANN system. A newly constructed recurrent AANN (R-AANN) was thus obtained by combining together the characteristics associated with AANN and RNN. This R-AANN validity was first demonstrated using model problems consisting of a continuous spring-mass linear oscillator and a set of nonlinear discrete recurrence relationships. Longitudinal flight simulation was then conducted to generate the training and testing flight data. It was shown that both static AANN and dynamic R-AANN yield good data compatibility check ability of the parameters selected in the input node layer. However, R-AANN performance is superior in the validation of the underlying dynamics in terms of a better representation of the time derivatives of the flight parameters.

[1]Klein, V., “Estimation of Aircraft Aerodynamics Parameters from Flight Data,” Progress in Aerospace Sciences, Vol. 26, 1989, pp. 1-77.
[2]Iliff, K. W., “Parameter Estimation for Flight Vehicles,” Journal of Guidance, Control, and Dynamics, Vol. 12, No. 5, 1989, pp. 609-622.
[3]Laban, M., “On-Line Aircraft Aerodynamic Model Identification,” Proefschrift Technische Universiteit Delft.-Met Lit. Opg., 1994.
[4]詹銘堯, “非線性飛行動力系統參數識別與飛行軌跡重建,” 國立成功大學航空太空工程研究所碩士論文, 1997年.
[5]徐伯熊, “基因演算於飛行軌跡重建之應用,” 國立成功大學航空太空工程研究所碩士論文, 1999年.
[6]Napolitano, M. R., Neppach, C., Casdorph, V., and Naylor, S., “Sensor Failure Detection, Identification, and Accommodation Using On-Line Learning Neural Architectures,” AIAA Paper 94-3598, 1994.
[7]王紀瑞, “類神經網路服役負載估算及記錄器參數驗證與處理,” 國立成功大學航空太空工程研究所博士論文, 2003年.
[8]Lippman, R. P., “An Introduction to Computing with Neural Nets,” IEEE ASSP Magazine, April 1987, pp. 4-22.
[9]Haykin, S., Neural Networks, A Comprehensive Foundation, Macmillian College Publishing Co. Inc., New York, 1994.
[10]Fausett, L., Fundamentals of Neural Networks: Architectures, Algorithms and Applications, Prentice Hall, Englewood Cliffs, New Jersey, ISBN 0-13-334186-0, 1994.
[11]McLean, D., Automatic Flight Control System, Prentice Hall Inc., 1990.
[12]Nelson, R. C., Flight Stability and Automatic Flight Control, McGraw-Hill Inc., 1989.
[13]Etkin, B., Dynamics of Flight Stability and Control, John Wiley & Sons, Inc., 1982.
[14]Boiffier, J. L., The Dynamics of Flight The Equations, John Wiley & Sons Ltd, 1998.
[15]Stevens, B. L., and Lewis, F. L., Aircraft Control and Simulation, John Wiley & Sons, Inc., 1992.
[16]Rosenblatt, F., “The Perceptron : A Probabilistic Model for Information Storage and Organization in the Brain,” Psychological Review, Vol. 65, 1958, pp. 386-408.
[17]Rumelhart, D. E., Hinton, G. E., and Williams, R. J., “Learning Internal Representations by Error Propagation,” Parallel Distributed Processing, Chapter 8, MIT Press, 1986.
[18]馮煥清, 巢洪斌, “BP網路及其應用,” 中國科技大學生物醫學工程委員會, 儒林圖書股份有限公司, 台北市, 1993年1月, pp. 390-410.
[19]王煦法, 張作生, 巢洪斌, “神經網路在自動控制中的應用,” 中國科技大學生物醫學工程委員會, 儒林圖書股份有限公司, 台北市, 1993年1月, pp. 412-416.
[20]Kramer, M. A., “Autoassociative Neural Networks,” Computers and Chemical Engineering, Vol. 16, No. 4, 1992, pp. 313-328.
[21]Kramer, M. A., “Non-linear Principal Component Analysis Using Autoassociative Networks,” AICHE Journal, Vol. 37, No. 2, 1991, pp. 233-243.
[22]Napolitano, M., Windon, D., Casanova, J., and Innocenti, M., “A Comparison between Kalman Filter and Neural Network Approaches for Sensor Validation,” AIAA Paper 96-3894, 1996.
[23]Mattern, D. L., Jaw, L. C., Guo, T. H., Graham, R., and McCoy, W., “Using Neural Networks for Sensor Validation,” AIAA Paper 98-3547, 1998.
[24]Mattern, D. L., Jaw, L. C., Guo, T. H., Graham, R., and McCoy, W., “Simulation of an Engine Sensor Validation Scheme Using an Autoassociative Neural Networks,” AIAA Paper 97-2902, 1997.
[25]Kramer, M. A., “Nonlinear Principal Component Analysis Using Autoassociative Neural Networks,” Laboratory for Intelligent Systems in Process Engineering, Dept. of Chemical Engineering, Massachusetts Institute of Technology, Cambridge, MA 02139.
[26]Narendra, K. S., and Parthasarathy, K., “Identification and Control of Dynamical Systems Using Neural Networks,” IEEE Transactions on Neural Networks, Vol. 1, No. 1, March 1990, pp. 4-27.