Uçak Motoru Sağlığı için Uzun-Kısa Süreli Bellek Yöntemi ile Öngörücü Bakım

Year 2019, , 103 - 109, 30.04.2019

Merve Ayyüce Kızrak , Bülent Bolat

https://doi.org/10.17671/gazibtd.495730

Cited By: 5

Abstract

Otomotiv, uçak ve fabrika otomasyonu gibi özellikle maliyetli motor bakımı gerektiren uygulamalarda öngörücü bakım önemli bir yer almaktadır. Hem iş güvenliği hem de araçlardan sağlanacak verim bakımından motorların bakım periyotlarını önceden kestirmek ve buna göre iş yönetim stratejisi geliştirmek önemlidir. Öngörücü bakım için motorlardan alınan sensör verileri motorun yıpranma süresini ve seviyesini belirlemekte kullanılmaktadır. Çalışmada Uzun-Kısa Süreli Bellek (LSTM) yapısı kullanılarak, uçak motorlarının kalan yaşam ömürlerinin tahmin edilmesi amaçlanmıştır. NASA tarafından sunulmuş olan bir veri kümesi üzerinde LSTM yapısı test edilmiştir ve elde edilen sonuçlar farklı yöntemlerle kıyaslanmıştır. Yapılan uygulamaların sonucunda en yüksek sınıflandırma başarımı %98,876; en düşük ortalama mutlak hata ise 1,343 olarak LSTM ile elde edilmiştir.

Keywords

öngörücü bakım, derin öğrenme, uzun-kısa vadeli bellek, makine öğrenmesi, özyinelemeli sinir ağları, uçak motor sağlığı, yapay sinir ağları

Predictive Maintenance of Aircraft Motor Health with Long-Short Term Memory Method

Abstract

Predictive maintenance is an important part of applications that require costly engine maintenance, such as automotive, aircraft and factory automation. It is important to anticipate the maintenance periods of the engines and develop a business management strategy accordingly in terms of both work safety and efficiency. For predictive maintenance, the sensor data from the motors were used to determine the wear time and level of the engine. In this study, a solution based on deep learning is proposed as an alternative to traditional regression and classification methods. The NASA Turbofan Engine Corruption Simulation data set was studied by using Long-Short Term Memory (LSTM), one of the deep learning models and known to make successful predictions on time-dependent data such as time series. During the simulations, the highest classification performance and the lowest mean absolute error were obtained by LSTM as 98,876% 1.343 respectively.

Keywords

predictive maintenance, deep learning, long-short term memory, artificial neural networks, machine learning, aircraft health system

References

• [1] V. Fornlöf, Improved Remaining Useful Life Estimations for On-Condition Parts in Aircraft Engines, University of Skövde, Sweden, Runit AB, Skövde, ISBN 978-91-981474-9-0, 2016.
• [2] A. Kumar, R. Shankar, L. S. Thakur, “A Big Data Driven Sustainable Manufacturing Framework for Condition-Based Maintenance Prediction”, Journal of Computational Science, Elsevier B.V., (27) 2018, 2018.
• [3] J. Xu, Y. Wang, Y. Xu, “PHM-Oriented Integrated Fusion Prognostics for Aircraft Engines Based on Sensor Data”, IEEE Sensors Journal, 14(4), 2014.
• [4] P. Malhotra, A. Ramakrishnan, G. Anand, L. Vig, P. Agarwal, G. Shroff, “LSTM-based Encoder-Decoder for Multi-sensor Anomaly Detection”, Presented at ICML 2016 Anomaly Detection Workshop, New York, NY, USA, 2016.
• [5] E. Ramasso, A. Saxena, “Performance Benchmarking and Analysis of Prognostic Methods for CMAPSS Datasets”, International Journal of Prognostics and Health Management, ISSN 2153-2648, 014, 2014.
• [6] X. Zhang, L. Tang, “Decastro J., Robust Fault Diagnosis of Aircraft Engines: A Nonlinear Adaptive Estimation-Based Approach”, IEEE Transactions on Control Systems Technology, 21(3), 2013.
• [7] G. S. Babu, P. Zhao, X-L. Li, “Deep Convolutional Neural Network Based Regression Approach for Estimation of Remaining Useful Life”, DASFAA: Database Systems for Advanced Applications, 214-228, 2016.
• [8] C. Zhang, J. H. Sun, K. C. Tan, “Deep Belief Networks Ensemble with Multi-objective Optimization for Failure Diagnosis”, IEEE International Conference on Systems, Man, and Cybernetics, 32-37, 2015.
• [9] Y. Lei, N. Li, L. Guo, N. Li, T. Yan, J. Lin, “Machinery health prognostics: A systematic review from data acquisition to RUL prediction”, Mechanical Systems and Signal Processing, Elsevier Ltd, 104, 799–834, 2017.
• [10] Internet: A. Saxena, K. Goebel, Turbofan Engine Degradation Simulation Data Set, NASA The Prognostics Data Repository, http://ti.arc.nasa.gov/project/prognostic-data-repository, NASA Ames Research Center, Moffett Field, CA, 01.05.2018.
• [11] W. Zhang, C. Li, G. Peng, Y. Chen, Z. Zhang, “A Deep Convolutional Neural Network with New Training Methods for Bearing Fault Diagnosis Under Noisy Environment and Different Working Load”, Mechanical Systems and Signal Processing, 100, 439–453, 2018.
• [12] J. Chung, C. Gulcehre, K. Cho, Y. Bengio, “Empirical Evaluation of Gated Recurrent Neural Networks on Sequence Modeling”, Advances in Neural Information Processing Systems (NIPS), Deep Learning and Representation Learning Workshop, 2014.
• [13] S. Khan, T. Yairi, “A review on the Application of Deep Learning in System Health Management”, Elsevier Journal of Mechanical Systems and Signal Processing, 107, 241–265, 2018.
• [14] S. Hochreiter, J. Schmidhuber, “Long Short-Term Memory”, Neural Computation, 9(8), 1735-1780, 1997.
• [15] P. K. Diederik, B. Jimmy, “ADAM: A Method for Stochastic Optimization”, 3rd International Conference for Learning Representations, San Diego, 2014.
• [16] M. Yuan, Y. Wu, L. Lin, “Fault Diagnosis and Remaining Useful Life Estimation of Aero Engine Using LSTM Neural Network”, IEEE International Conference on Aircraft Utility Systems (AUS), 135–140, 2016.
• [17] N. Gugulothu, V. Tv, P. Malhotra, L. Vig, P. Agarwal, G. Shroff, “Predicting Remaining Useful Life Using Time Series Embeddings Based on Recurrent Neural Networks”, 2nd ML for PHM Workshop at SIGKDD, Halifax, Canada, preprint arXiv:1709.01073, 2017.
• [18] P. Malhotra, A. Ramakrishnan, G. Anand, L. Vig, P. Agarwal, G. Shroff, “Multi-sensor Prognostics Using an Unsupervised Health Index Based on LSTM Encoder-Decoder”, 1st ACM SIGKDD Workshop on Machine Learning for Prognostics and Health Management, USA, preprint arXiv:1608.06154, 2016.
• [19] Y. Wu, M. Yuan, S. Dang, L. Lin, Y. Liu, “Remaining Useful Life Estimation of Engineered System Using Vanilla LSTM Neural Networks”, Journal of Neurocomputing, Elsevier Pub., 275, 167-179, 2018.
• [20] R. Zhao, J. Wang, R. Yan, K. Mao, “Machine Health Monitoring with LSTM Networks”, Tenth International Conference on Sensing Technology, 2016.
• [21] R. Zhao, R. Yan, J. Wang, K. Mao, “Learning to Monitor Machine Health with Convolutional Bi-Directional LSTM Networks”, Sensors (Basel), (2), 273, doi: 10.3390/s17020273, 2017.
• [22] Internet: M. A. Kızrak, “Uçak Motoru Sağlığı için Uzun-Kısa Süreli Bellek Yöntemi ile Öngörücü Bakım”, GitHub Repo kaynak kodları, https://github.com/ayyucekizrak, 30.11.2018.