Single and Multiple Failures Diagnostics of Pneumatic Valves using Machine Learning

• Humberto Sano Universidade Federal de São Paulo
• João Malere Instituto Tecnológico de Aeronáutica
• Lilian Berton Universidade Federal de São Paulo

DOI: https://doi.org/10.5753/eniac.2019.9284

Resumo

Predictive maintenance of aeronautical systems is an important field of study to reduce airlines operational costs and non-scheduled events. Valves are important components of several aircraft systems. This fact motivates the need for health monitoring and prognosis. In this context, this work compares machine learning methods to predict pneumatic valves health conditions. A multilayer perceptron artificial neural network model was able to discriminate the three single failures modes with a mean accuracy of 99.9% and a support vector machine model was able to diagnose single and concurrent failure modes with an average accuracy of 94.3%. The results reveal an accuracy improvement compared to a previous pneumatic valves health assessment study.

Palavras-chave: Machine Learning, Diagnostic, Maintenance, Concurrent Failure, Multi-Failure, Valves

Referências

Baptista, M., Nascimento Jr, C., Prendinger, H., and Henriques, E. (2017a). A case for the use of data-driven methods in gas turbine prognostics. In Annual Conference of the Prognostics and Health Management Society, pages 1-10.

Baptista, M., P. de Medeiros, I., Malere, J., Nascimento Jr, C., Prendinger, H., and Henriques, E. (2017b). Comparative case study of life usage and data-driven prognostics techniques using aircraft fault messages. Computers in Industry, 86:1–14.

Caesarendra, W., Widodo, A., and Yang, B.-S. (2010). Application of relevance vector machine and logistic regression for machine degradation assessment. Mechanical Systems and Signal Processing, 24(4):1161 – 1171.

Castilho, H. M., Nascimento, C. L., and Vianna, W. O. L. (2018). Aircraft bleed valve fault classification using support vector machines and classification trees. In 2018 Annual IEEE International Systems Conference (SysCon), pages 1–7.

Daigle, M. J. and Goebel, K. (2013). Model-based prognostics with concurrent damage progression processes. IEEE Transactions on Systems, Man, and Cybernetics: Systems, 43(3):535–546.

Dimitriadou, E., Hornik, K., Leisch, F., Meyer, D., Weingessel Maintainer, A., Leisch@ci, f., Tuwien, A., and , A. (2006). The e1071 package. pages 32-48.

Günther, F. and Fritsch, S. (2010). neuralnet: Training of neural networks. R Journal, 2:30–38.

Hastie, T., Tibshirani, R., and Friedman, J. (2009). The elements of statistical learning: data mining, inference and prediction. Springer, 2 edition.

Jennions, I. K. (2014). Integrated Vehicle Health Management: Implementation and Lessons Learned. SAE International.

M. Therneau, T. and Atkinson, E. (1997). An introduction to recursive partitioning using the rpart routines. Mayo Clinic, 61:5–9.

Mitchell, T. M. (1997). Machine Learning. McGraw-Hill, Inc., New York, NY, USA, 1 edition.

PHMSociety (2019). Phmsociety. https://www.phmsociety.org. Accessed: 2019-05-25.

Ripley, B. and Venable, W. (2019). Functions for classification. https://cran.r-project.org/web/packages/class/class.pdf. Accessed: 2019-0601.

Russell, S. and Norvig, P. (2009). Artificial Intelligence: A Modern Approach. Prentice Hall Press, Upper Saddle River, NJ, USA, 3rd edition.

Tibshirani, R., Hastie, T., and Friedman, J. (2010). Regularized paths for generalized linear models via coordinate descent. Journal of Statistical Software, 33:9–10.

Turcio, W., Yoneyama, T., and Moreira, F. (2013). Quasi-l6pv gain-scheduling control of a nonlinear aircraft pneumatic system. In 21st Mediterranean Conference on Control and Automation, pages 341–350.

Vianna, W. O. L. and Yoneyama, T. (2018). Predictive maintenance optimization for aircraft redundant systems subjected to multiple wear profiles. IEEE Systems Journal, 12(2):1170–1181.