本研究以基因演算自相關類神經網路(Genetic Auto-associative Artificial Neural Network, GA-AANN)在發動機傳感器訊號驗證的成果為基礎，將其擴展至即時(Real-time)訊號處理的應用上，發展一套更符合實際需求的工具。本研究運用分散式神經網路（Decentralized Neural Networks, DNN），結合雜訊濾除神經網路 (Noise-filtering AANN) 以及自映射神經網路 (Self-mapping AANN)，發展出具有即時性的發動機傳感器故障檢出程序，稱之為DNN-AANN，以改善目前大多數系統皆偏重於離線式的訊號驗證。本研究使用普惠（Pratt & Whitney）公司PW4000-94型發動機之影響係數矩陣（Influence Coefficient Matrix），產生網路訓練與測試所需之故障樣本。結果顯示，在量測參數充足（8個）的情況下，DNN-AANN網路可成功、即時地判斷出發動機故障的傳感器，並且以分散式神經網路所提供的一較為準確良好的初始猜測值供後續的基因演算法則使用，因此將可更加快速的逼近全域最佳解，以此進行即時訊號的識別與補正。

A real-time engine sensor data validation procedure is developed based on the previous framework of Genetic Auto-associative Artificial Neural Network (GA-AANN). The objective is to accelerate the processing speed of the off-line GA-AANN system by incorporating a bank of Decentralized Neural Networks (DNNs) into the existing GA-AANN procedure comprising Noise-filtering and Self-mapping networks. This improved procedure, termed DNN-AANN, was justified to be able to perform in a real-time manner the sensor validation tasks including sensor failure detection, identification, and accommodation (SFDIA). It was found in the present DNN-AANN procedure that DNN can quickly isolate the faulty sensor and provide a good initial guess for bias correction using a global genetic search algorithm. The training and testing gaspath samples were generated using an influence coefficient matrix of the PW 4000-94 engine, augmented by a noise and bias model. This newly developed SFDIA procedure can only work with extended systems with sufficient amount of sensors. For limited systems, the data redundancy is not enough for constructing the DNN bank which are the constituent elements of the present algorithm. This new real-time SFDIA alogrithm has been successfully proven using 66000 samples covering 15 fault types. The robustness is also verified, and in additoin, the influences of sensor noise and bias on the validation results were examined.

[1] Andrenucci, M., Lazzeretti, R., “Problems in Fault Diagnostics and Prognostics for Engine Condition Monitoring,” AGARD Conference Proceedings, No. 165 Part A, 4-5 April, 1974.
[2] 陳大光, “發動機狀態監視及故障診斷系統的經濟效益分析,” 幹線客機用發動機狀態監視及故障診斷系統的分析研究, 北京航空航天大學動力系, 1993年5月.
[3] 陳大光, 張赫然, “用氣動熱力參數對發動機故障進行診斷與定位的實用方法,” 北京航空航天大學碩士論文第二章, 1989年.
[4] Urban, L. A., “Gas Path Analysis Applied to Turbine Engine Condition Monitoring,” AIAA 72-1082, 1972.
[5] Urban, L. A., and Volponi, Allan J., “Mathematical Methods of Relative Engine Performance Diagnostics,” Journal of Aerospace, Vol. 101, Technical Paper 922048, 1992, pp. 2025-2049.
[6] Doel, D. L., “TEMPER-A Gas Path Analysis Tool for Commercial Jet Engines,” ASME Journal of Engineering for Gas Turbines and Power, Vol.116, No. 1, 1994, pp. 82-89.
[7] Gelb, A. ed., “Applied Optimal Estimation,” The MIT Press, Cambridge, MA, pp. 102-103, 1974.
[8] Doel, D.L., “An Assessment of Weighted-Least-Squares-Based Gas Path Analysis,” ASME Journal of Engineering for Gas Turbines and Power, Vol. 116, April 1994, pp. 366-373.
[9] Rosenblatt, F., “The Perceptron: A Probabilistic Model for Information Storage and Organization in the Brain,” Psychological Review 65, 1958, pp. 386-408.
[10] Haykin, S., “Neural Networks, A Comprehensive Foundation,” Macmillan College Publishing Co. Inc., 1994, New York.
[11] Fausett, L., “Fundamentals of Neural Networks: Architectures, Algorithms and Applications,” Prentice Hall, Englewood Cliffs, New Jersey, ISBN 0-13-334186-0, 1994.
[12] Eustace, R., and Merrington, G., “Fault Diagnosis of Fleet Engines Using Neural Networks,” ISABE 95-7085, 1995, pp.926-936.
[13] Cifald, M. L., and Chokani, N., “Engine Monitoring Using Neural Networks,” AIAA Paper 98-3548, 1998.
[14] Zedda, M., and Singh, R., “Fault Diagnosis of a Turbine Engine Using Neural Networks: a Quantitative Approach,” AIAA Paper 98-3602, 1998.
[15] Luce, H., and Govind, R., “Neural Nerwork Pattern Recognizer for Detection of Failure Modes in the SSME,” AIAA Paper 90-1893, 1990.
[16] Denney, G., “F-16 Jet Engine Trending and Diagnostics with Neural Networks,” Applications of Neural Networks Ⅳ, SPIE Vol. 1965, pp. 419-422.
[17] Torella, G., Lombardo, G., “Utilization of Neural Networks for Gas Turbine Engines,” ISABE 95-7032, Italian Air Force Academy, 1995.
[18] 陳大光, 劉福生, “燃氣渦輪發動機故障診斷的人工神經網絡法,” CSAA 93-P-184(P), 北京航空航天大學動力系, 1993年5月.
[19] Eustace, R., “Neural Network Fault Diagnosis of A Turbofan Engine,” Proceedings of the International Symposium on Air Breathing Engines, Vol. 2, Tokyo, Japan, Sept, 20-24, 1993.
[20] 陸鵬舉, 徐自珍, “航空發動機人工神經網路故障診斷法,” 中國航空太空學會學刊, Vol. 32, No. 2, December 2000, pp. 301-308.
[21] P. J. Lu, M. C. Zhang, T. C. Hsu and J. Zhang, “An Evaluation of Engine Faults Diagnostics Using Artificial Neural Networks,” ASME Journal of Engineering for Gas Turbines and Power, Vol. 123, No. 2, April 2001, pp. 340-346.
[22] T. C. Hsu, “Engine Gaspath Sensor Data Validation and Fault Diagnosis Using Artificial Neural Networks,” Ph.D. Thesis, Institute of Aeronautics and Astronautics, National Cheng Kung University, 2002.
[23] P. J. Lu and T. C. Hsu, “Application of Autoassociative Neural Network on Gaspath Sensor Data Validation,” AIAA Journal of Propulsion and Power, Vol. 18, No. 4, July-Aug., 2002, pp. 879-888.
[24] Kramer, M. A., “Autoassociative Neural Networks,” Computers and Chemical Engineering, Vol. 16, No. 4, 1992, pp. 313-328.
[25] Kramer, M. A., “Non-linear Principal Component Analysis Using Autoassociative Networks,” AICHE Journal, Vol. 37, No. 2, 1991, pp. 233-243.
[26] Napolitano, M., Windon, D., Casanova, J., and Innocenti, M., “A Comparison between Kalmam Filter and Neural Network Approaches for Sensor Validation,” AIAA Paper 96-3894, 1996.
[27] 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.
[28] 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.
[29] 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.
[30] Holland, J. H., Concerning Efficient Adaptive Systems, In M. C. Yovits, G. T. Jacobi, and G. D. Goldstein (Eds.), Self-Organizing System, Washington: Spartan Books. 1962, pp. 215-230.
[31] B. D. MacIsaac, “Engine Performance and Health Monitoring Models Using Steady State and Transient Prediction Methods,” AGARD Lecture Series 183, May 1992, pp. 9-1~9-21.
[32] Hagan M. T. and Menhaj M. B., “Training Feedforward Networks with the Marquardt Algorithm,” IEEE Transactions on Neural Networks, vol. 5, no. 6, November 1994, pp. 989-993.