Abstract
Neural networks are being tested to monitor aircraft engine condition data, in addition to current techniques and newer methods such as knowledge-based systems or case-based reasoning, in order to increase safety and assist in aircraft maintenance activity. It is possible that neural networks can help to measure subtle changes in a wide number of variables, and produce indications of adverse trends to serve as early warning signals. Unsupervised networks were trained on 300 records, each with 31 attributes, and independently validated on 1662 records (the recall set). Results are presented for self-organising maps and recirculation networks. The next phase is to incorporate diagnostic capability by adding a supervised learning element. Monitoring of sensor reliability and the provision of confidence limits are further extensions of these approaches.
Similar content being viewed by others
References
Kerry P, Page M, Ward MJ. COMPASS Analysis Guide for RB211-524G-EMMU Engine. Rolls-Royce, 1988–91
Hertz J, Krogh A, Palmer R. Introduction to the Theory of Neural Computation. New York. Addison-Wesley, 1991
Kohonen T. Self-Organization and Associative Memory. Springer-Verlag, Berlin Heidelberg New York, 1989
Durbin R, Willshaw D. An analogue approach to the travelling salesman problem using an elastic net method. Nature 1987; 326: 689–691
Frean M. The upstart algorithm: A method for constructing and training feedforward neural networks. Neural Computation 1990; 2: 198–209
Rumelhart DE, Hinton G, Williams R. Learning internal representations by error propagation. In: Rumelhart DE, McClelland JL, editors. Parallel Distributed Processing (Vol I). Cambridge, MA: MIT Press, 1986
Hinton G, McClelland JL. Learning representations by recirculation. In: Proceedings of IEEE Conference: Advances in Neural Information Processing Systems, San Mateo, CA: Morgan Kaufmann, 1988
Carpenter, Grossberg. The ART of adaptive pattern recognition by a self-organizing neural network. Computer 1988; 77–88, Mar
Dodd N. Intensive care ward monitoring and default training. In: Proceedings of INNC, Vol. 1; 1990; Paris. New York: Kluwer Academic, 1990: 326–329
Mathis D. Electronic mail communication to newsgroup comp.ai.neural-nets, 1991
Gallant S. Optimal linear discriminants. In: Proceedings 8th IEEE Conference on Pattern Recognition; 1986; Paris
Mezard M, Nadal JP. Learning in feedforward layered networks: The tiling algorithm. Journal of Physics ‘A’ 1989; 22: 2191–2203
Wynne-Jones M. Node splitting: A constructive algorithm for feed-forward neural networks. Neural Computing & Applications 1993; 1: 17–22
Wynne-Jones M. Node splitting: A constructive algorithm for feed-forward neural networks. In: Moody JE, Hanson SJ, Lippman RP (eds). Advances in neural information processing systems, 4, Morgan-Kaufmann, San Mateo, CA, 1992
Sietsma J, Dow R. Neural net pruning — why and how. In: Proceedings of IEEE International Conference on Neural Networks, Vol. 1; 1988; San Diego, CA: 325–333
Sietsma J, Dow R. Creating artificial neural networks that generalize. Neural Networks 1991; 4: 67–79
Hirose, Yamashita, Hijiya. Back-propagation algorithm which varies the number of hidden units. Neural Networks 1991; 4: 61–66
Specht DF. Probabilistic neural networks for classification, mapping or associative memory. In: Proceedings of the Conference on Neural Networks; 1988
Specht DF. Probabilistic neural networks. Neural Networks 1990; 3 (1): 109–118
Neural Ware, Inc. Neural Computing. Neural Works Professional II Plus manual, 1991
MacKay D. Bayesian Methods for Adaptive Models [dissertation]. Los Angeles, CA: CalTech, 1991
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
Cumming, S. Neural networks for monitoring of engine condition data. Neural Comput & Applic 1, 96–102 (1993). https://doi.org/10.1007/BF01411378
Received:
Issue Date:
DOI: https://doi.org/10.1007/BF01411378