Araştırma Makalesi
BibTex RIS Kaynak Göster

Derin Öğrenme Modelleri Kullanarak Endüstriyel Makinelerin Kalan Kullanım Ömrü Tahmini

Yıl 2025, Cilt: 13 Sayı: 3
https://doi.org/10.29109/gujsc.1598534

Öz

Bu çalışmada, endüstriyel makinelerde arıza tahmini için derin öğrenme tekniklerinin etkinliği araştırılmıştır. Araştırmada, olası makine arızalarını önceden tahmin etmek amacıyla sensör verileri üzerinde CNN, LSTM ve BiLSTM gibi farklı derin öğrenme mimarileri kullanılarak kombinasyonel modeller oluşturulmuştur. Farklı derin öğrenme mimarilerinin katman sayıları ve performansları ayrıntılı olarak incelenmiş, özellikle BiLSTM ve LSTM mimarilerinin zaman serisi veri analizi konusunda başarılı olduğu bulunmuştur. Sonuçlar, CNN-3BiLSTM yönteminin geleneksel ve makine öğrenimi yaklaşımlarına kıyasla daha düşük hata oranları ile başarılı sonuçlar verdiğini göstermiştir. Literatürdeki benzer çalışmalarla karşılaştırıldığında, önerilen yöntemin Çok Katmanlı Algılayıcı, Destek Vektör Regresyonu ve CNN gibi yöntemlere kıyasla önemli bir performans üstünlüğü sergilediği görülmüştür. Bu çalışma, derin öğrenme tabanlı arıza tahmin modellerinin endüstriyel makinelerde güvenliği artırma ve bakım maliyetlerini optimize etme konusundaki önemli potansiyelini ortaya koymaktadır.

Kaynakça

  • [1] Yong, Y., Stecke, K. E., Li, D., The evolution of production systems from Industry 2.0 through Industry 4.0. International Journal of Production Research. 2018; 56: 848-861.
  • [2] Zizic, M. C., Mladineo, M., Gjeldum, N., Celent, L., From industry 4.0 towards industry 5.0: A review and analysis of paradigm shift for the people, organization and technology. Energies. 2022; 15: 1-20.
  • [3] Yaman, G., Karadayı, H. M., Titreşim analizi ile pompalarda arıza tespiti ve kestirimci bakım için örnek bir çalışma. Tesisat Mühendisliği. 2014; 140: 37-51.
  • [4] Bektaş, O., Kestirimci bakım için döner mekanizma bozulma eğrisinin tanımlanması. Avrupa Bilim ve Teknoloji Dergisi. 2020; 19: 420-428.
  • [5] Günindi, Z. Ö., Kurban, A. O., Danışmaz, M., Soğutma fanının kestirimci bakım yöntemi uygulaması ile titreşim analizi ve yapısal esneklik arızasının giderilmesi. Icontech International Journal. 2022; 6: 33-41.
  • [6] Demir, H. G., Müştak, O., Rulman hasarlarının titreşim ve gürültü analizi ile tespiti. Avrupa Bilim ve Teknoloji Dergisi. 2021; 25: 571-581.
  • [7] Orhan, S., Arslan, H., Aktürk, N., Titreşim analiziyle rulman arızalarının belirlenmesi. Gazi Üniversitesi Mühendislik Mimarlık Fakültesi Dergisi. 2003; 18: 39-48.
  • [8] Özkat, E. C., Makine öğrenmesi metodolojisi kullanılarak yüksek hızlı rulmanlarda sağlık göstergesinin belirlenmesi. Avrupa Bilim ve Teknoloji Dergisi. 2021; 22: 176-183.
  • [9] Calayır, G. N., Kabak, M., Bakım için makine öğrenme tekniklerinin analizi ve bir uygulama. Journal of Turkish Operations Management. 2021; 5: 662-675.
  • [10] Güler, O., Turbofan motorlarının kestirimci bakımında makine öğrenimi algoritmaları performanslarının karşılaştırılması. Niğde Ömer Halisdemir Üniversitesi Mühendislik Bilimleri Dergisi. 2024; 13: 99-106.
  • [11] Görgülü, Ş., Koç, İ., Kırımça, N., Karaköse, M., vd., Yapay zekâ kullanımıyla peron ayırıcı kapı sisteminin sağlığını izleme ve kestirimci bakım. Akıllı Ulaşım Sistemleri ve Uygulamaları Dergisi. 2024; 7: 56-70.
  • [12] Ahmad, W., Khan, S. A., Islam, M. M., Kim, J. M., A reliable technique for remaining useful life estimation of rolling element bearings using dynamic regression models. Reliability Engineering & System Safety. 2019; 184: 67-76.
  • [13] Costello, J. J. A., West, G. M., McArthur, S. D. J., Machine learning model for event-based prognostics in gas circulator condition monitoring. IEEE Transactions on Reliability. 2017; 66: 1048-1057.
  • [14] Bukhsh, Z. A., Saeed, A., Stipanovic, I., Doree, A. G., Predictive maintenance using tree-based classification techniques: A case of railway switches. Transportation Research Part C: Emerging Technologies. 2019; 101: 35-54.
  • [15] Prytz, R., Nowaczyk, S., Rögnvaldsson, T., Byttner, S., Predicting the need for vehicle compressor repairs using maintenance records and logged vehicle data. Engineering Applications of Artificial Intelligence. 2015; 41: 139-150.
  • [16] Saimurugan , M., Ramachandran, K. I., Sugumaran, V., Sakthivel, N. R., Multi component fault diagnosis of rotational mechanical system based on decision tree and support vector machine. Expert Systems with Applications. 2011; 38: 3819-3826.
  • [17] Hsu, J. Y., Wang, Y. F., Lin, K. C., Chen, M. Y., Hsu, J. H. Y., Wind turbine fault diagnosis and predictive maintenance through statistical process control and machine learning. IEEE Access. 2020; 8: 23427-23439.
  • [18] Baştürk, M. Ö., Yüksel, M., Tetik, Y. E., Yılmaz, M., Güner, M., Kaya, T., Pantograf boynuz hatalarının derin öğrenme ve görüntü işleme teknikleri ile tespiti. Demiryolu Mühendisliği. 2022; 16: 102-115.
  • [19] Emre, Y., Özmen, N. G., Rulmanlarda titreşim verilerinden durum izleme ve arıza teşhisi için derin öğrenme yönteminin uygulanması. Konya Journal of Engineering Sciences. 2022; 10: 346-365.
  • [20] Çatak, Y., Şahin, K., Güney, O. B., Özkan, H., Deep convolutional autoencoder architecture for predictive maintenance applications. 30th Signal Processing and Communications Applications Conference (SIU). 2022; 1-4.
  • [21] De Santo, A., Ferraro, A., Galli, A., Moscato, V., Sperlì, G., Evaluating time series encoding techniques for predictive maintenance. Expert Systems with Applications. 2022; 210: 118435.
  • [22] Wang, Y., Long, M., Wang, J., Gao, Z., Yu, P. S., Predrnn: Recurrent neural networks for predictive learning using spatiotemporal lstms. Advances in Neural Information Processing Systems. 2017; 30.
  • [23] Wang, W., Zhou, G., He, Z., Zhou, P., Ma, G., Yan, X., Ma, T., A novel unsupervised adaptive density-based clustering filter for remaining useful life prediction of bearings. Measurement Science and Technology. 2024; 35: 126121.
  • [24] Wang, H., Ding, X., Huang, Z., Zhen, H., Yu, W., Huang, W., Multi-stage exponential model based on subspace clustering distribution for bearing remaining useful life prediction. Measurement Science and Technology. 2024; 35: 076129.
  • [25] Adnan, A., Faris, E., Amaral, T. J., Fault diagnosis and health management of bearings in rotating equipment based on vibration analysis - a review. Journal Of Vibroengineering. 2022; 24: 46-74.
  • [26] Jianghong, Z., Jiahong, Y., Quan, Q., Yi, Q., A comprehensive survey of machine remaining useful life prediction approaches based on pattern recognition: taxonomy and challenges. Measurement Scıence And Technology. 2024; 35: 062001.
  • [27] Xianpeng, Q., Veronica, J., Seong, L., Chin L., Tiyamike, B., Advances and limitations in machine learning approaches applied to remaining useful life predictions: a critical review. Internatıonal Journal Of Advanced Manufacturıng Technology. 2024; 133: 4059-4076.
  • [28] Salman, K., Soo-Ho, J., Yaseen, S. S., Ha, J. J., Soo, K. H., Artificial Intelligence-Driven Prognostics and Health Management for Centrifugal Pumps: A Comprehensive Review. Actuators. 2024; 13: 514.
  • [29] Luo, J., Shao, H., Lin, J., Liu, B., Meta-learning with elastic prototypical network for fault transfer diagnosis of bearings under unstable speeds. Reliability Engineering & System Safety. 2024; 245: 110001.
  • [30] Lei, Z., Su, Y., Feng, K., Wen, G., Interpretable operational condition attention-informed domain adaptation network for remaining useful life prediction under variable operational conditions. Control Engineering Practice. 2024; 153: 106080.
  • [31] Phung, V. H., Rhee, E. J., A High-Accuracy Model Average Ensemble of Convolutional Neural Networks for Classification of Cloud Image Patches on Small Datasets. Applied Sciences. 2019; 9: 4500.
  • [32] Akay, B., Karaboga, D., Akay, R., A comprehensive survey on optimizing deep learning models by metaheuristics. Artificial Intelligence Review. 2022; 55: 829-894.
  • [33] Zhang, X., Shi, J., Zhu, G., Xiao, Y., Chen, H., Study of regional monthly precipitation based on CEEMD-BILSTM coupled model. Water Supply. 2022; 22: 8036-8049.
  • [34] Dey, R., Salem, F. M., Gate-variants of gated recurrent unit (GRU) neural networks. IEEE 60th International Midwest Symposium on Circuits and Systems. 2017; 1597-1600.
  • [35] Veloso, G. A., Loschi, R. H., Dynamic linear degradation model: Dealing with heterogeneity in degradation paths. Reliability Engineering & System Safety. 2021; 210: 107446.
  • [36] Frederick, D. K., DeCastro, J. A., Litt, J. S. (2007). User's Guide for the Commercial Modular Aero-Propulsion System Simulation (C MAPSS). NASA/ARL, Technical Manual TM2007-215026.
  • [37] Aydın, H., Effects of air to fuel ratio on parameters of combustor used for gas turbine engines: Applications of turbojet, turbofan, turboprop and turboshaft. Energy. 2024; 305: 132346.
  • [38] Babu, G. S., Zhao, P., Li, X. L., Deep convolutional neural network based regression approach for estimation of remaining useful life. International Conference on Database Systems for Advanced Applications. Springer, Cham. 2016; 214-228.
  • [39] Zheng, S., Ristovski, K., Farahat, A., Gupta, C., Long short-term memory network for remaining useful life estimation. 2017 IEEE International Conference on Prognostics and Health Management (ICPHM). 2017.
  • [40] Zhang, J., Wang, P., Yan, R., Gao, R. X., Long short-term memory for machine remaining life prediction. Journal of Manufacturing Systems. 2018; 48: 78-86.
  • [41] Azyus, A. F., Wijaya, S. K., Naved, M., Prediction of remaining useful life using the CNN-GRU network: A study on maintenance management. Software Impacts. 2023; 17: 100535

Remaining Useful Life Prediction of Industrial Machines using Deep Learning Models

Yıl 2025, Cilt: 13 Sayı: 3
https://doi.org/10.29109/gujsc.1598534

Öz

In this study, the effectiveness of deep learning techniques in fault prediction for industrial machines was investigated. The research employed combinational models created using different deep learning architectures like CNN, LSTM, and BiLSTM on sensor data to predict potential machine failures in advance. Additionally, the number of layers and performances of different deep learning architectures were examined in detail, with BiLSTM and LSTM architectures being found especially successful in time series data analysis. As a result, the CNN-3BiLSTM method particularly demonstrated successful outcomes with lower error rates compared to traditional and machine learning approaches, thereby demonstrating the applicability of deep learning models in predictive maintenance applications. When compared with similar studies in the literature, the proposed method exhibited significant performance superiority over methods such as Multilayer Perceptron, Support Vector Regression, and CNN. The study revealed the substantial potential of deep learning-based fault prediction models in enhancing industrial machine safety and optimizing maintenance costs.

Kaynakça

  • [1] Yong, Y., Stecke, K. E., Li, D., The evolution of production systems from Industry 2.0 through Industry 4.0. International Journal of Production Research. 2018; 56: 848-861.
  • [2] Zizic, M. C., Mladineo, M., Gjeldum, N., Celent, L., From industry 4.0 towards industry 5.0: A review and analysis of paradigm shift for the people, organization and technology. Energies. 2022; 15: 1-20.
  • [3] Yaman, G., Karadayı, H. M., Titreşim analizi ile pompalarda arıza tespiti ve kestirimci bakım için örnek bir çalışma. Tesisat Mühendisliği. 2014; 140: 37-51.
  • [4] Bektaş, O., Kestirimci bakım için döner mekanizma bozulma eğrisinin tanımlanması. Avrupa Bilim ve Teknoloji Dergisi. 2020; 19: 420-428.
  • [5] Günindi, Z. Ö., Kurban, A. O., Danışmaz, M., Soğutma fanının kestirimci bakım yöntemi uygulaması ile titreşim analizi ve yapısal esneklik arızasının giderilmesi. Icontech International Journal. 2022; 6: 33-41.
  • [6] Demir, H. G., Müştak, O., Rulman hasarlarının titreşim ve gürültü analizi ile tespiti. Avrupa Bilim ve Teknoloji Dergisi. 2021; 25: 571-581.
  • [7] Orhan, S., Arslan, H., Aktürk, N., Titreşim analiziyle rulman arızalarının belirlenmesi. Gazi Üniversitesi Mühendislik Mimarlık Fakültesi Dergisi. 2003; 18: 39-48.
  • [8] Özkat, E. C., Makine öğrenmesi metodolojisi kullanılarak yüksek hızlı rulmanlarda sağlık göstergesinin belirlenmesi. Avrupa Bilim ve Teknoloji Dergisi. 2021; 22: 176-183.
  • [9] Calayır, G. N., Kabak, M., Bakım için makine öğrenme tekniklerinin analizi ve bir uygulama. Journal of Turkish Operations Management. 2021; 5: 662-675.
  • [10] Güler, O., Turbofan motorlarının kestirimci bakımında makine öğrenimi algoritmaları performanslarının karşılaştırılması. Niğde Ömer Halisdemir Üniversitesi Mühendislik Bilimleri Dergisi. 2024; 13: 99-106.
  • [11] Görgülü, Ş., Koç, İ., Kırımça, N., Karaköse, M., vd., Yapay zekâ kullanımıyla peron ayırıcı kapı sisteminin sağlığını izleme ve kestirimci bakım. Akıllı Ulaşım Sistemleri ve Uygulamaları Dergisi. 2024; 7: 56-70.
  • [12] Ahmad, W., Khan, S. A., Islam, M. M., Kim, J. M., A reliable technique for remaining useful life estimation of rolling element bearings using dynamic regression models. Reliability Engineering & System Safety. 2019; 184: 67-76.
  • [13] Costello, J. J. A., West, G. M., McArthur, S. D. J., Machine learning model for event-based prognostics in gas circulator condition monitoring. IEEE Transactions on Reliability. 2017; 66: 1048-1057.
  • [14] Bukhsh, Z. A., Saeed, A., Stipanovic, I., Doree, A. G., Predictive maintenance using tree-based classification techniques: A case of railway switches. Transportation Research Part C: Emerging Technologies. 2019; 101: 35-54.
  • [15] Prytz, R., Nowaczyk, S., Rögnvaldsson, T., Byttner, S., Predicting the need for vehicle compressor repairs using maintenance records and logged vehicle data. Engineering Applications of Artificial Intelligence. 2015; 41: 139-150.
  • [16] Saimurugan , M., Ramachandran, K. I., Sugumaran, V., Sakthivel, N. R., Multi component fault diagnosis of rotational mechanical system based on decision tree and support vector machine. Expert Systems with Applications. 2011; 38: 3819-3826.
  • [17] Hsu, J. Y., Wang, Y. F., Lin, K. C., Chen, M. Y., Hsu, J. H. Y., Wind turbine fault diagnosis and predictive maintenance through statistical process control and machine learning. IEEE Access. 2020; 8: 23427-23439.
  • [18] Baştürk, M. Ö., Yüksel, M., Tetik, Y. E., Yılmaz, M., Güner, M., Kaya, T., Pantograf boynuz hatalarının derin öğrenme ve görüntü işleme teknikleri ile tespiti. Demiryolu Mühendisliği. 2022; 16: 102-115.
  • [19] Emre, Y., Özmen, N. G., Rulmanlarda titreşim verilerinden durum izleme ve arıza teşhisi için derin öğrenme yönteminin uygulanması. Konya Journal of Engineering Sciences. 2022; 10: 346-365.
  • [20] Çatak, Y., Şahin, K., Güney, O. B., Özkan, H., Deep convolutional autoencoder architecture for predictive maintenance applications. 30th Signal Processing and Communications Applications Conference (SIU). 2022; 1-4.
  • [21] De Santo, A., Ferraro, A., Galli, A., Moscato, V., Sperlì, G., Evaluating time series encoding techniques for predictive maintenance. Expert Systems with Applications. 2022; 210: 118435.
  • [22] Wang, Y., Long, M., Wang, J., Gao, Z., Yu, P. S., Predrnn: Recurrent neural networks for predictive learning using spatiotemporal lstms. Advances in Neural Information Processing Systems. 2017; 30.
  • [23] Wang, W., Zhou, G., He, Z., Zhou, P., Ma, G., Yan, X., Ma, T., A novel unsupervised adaptive density-based clustering filter for remaining useful life prediction of bearings. Measurement Science and Technology. 2024; 35: 126121.
  • [24] Wang, H., Ding, X., Huang, Z., Zhen, H., Yu, W., Huang, W., Multi-stage exponential model based on subspace clustering distribution for bearing remaining useful life prediction. Measurement Science and Technology. 2024; 35: 076129.
  • [25] Adnan, A., Faris, E., Amaral, T. J., Fault diagnosis and health management of bearings in rotating equipment based on vibration analysis - a review. Journal Of Vibroengineering. 2022; 24: 46-74.
  • [26] Jianghong, Z., Jiahong, Y., Quan, Q., Yi, Q., A comprehensive survey of machine remaining useful life prediction approaches based on pattern recognition: taxonomy and challenges. Measurement Scıence And Technology. 2024; 35: 062001.
  • [27] Xianpeng, Q., Veronica, J., Seong, L., Chin L., Tiyamike, B., Advances and limitations in machine learning approaches applied to remaining useful life predictions: a critical review. Internatıonal Journal Of Advanced Manufacturıng Technology. 2024; 133: 4059-4076.
  • [28] Salman, K., Soo-Ho, J., Yaseen, S. S., Ha, J. J., Soo, K. H., Artificial Intelligence-Driven Prognostics and Health Management for Centrifugal Pumps: A Comprehensive Review. Actuators. 2024; 13: 514.
  • [29] Luo, J., Shao, H., Lin, J., Liu, B., Meta-learning with elastic prototypical network for fault transfer diagnosis of bearings under unstable speeds. Reliability Engineering & System Safety. 2024; 245: 110001.
  • [30] Lei, Z., Su, Y., Feng, K., Wen, G., Interpretable operational condition attention-informed domain adaptation network for remaining useful life prediction under variable operational conditions. Control Engineering Practice. 2024; 153: 106080.
  • [31] Phung, V. H., Rhee, E. J., A High-Accuracy Model Average Ensemble of Convolutional Neural Networks for Classification of Cloud Image Patches on Small Datasets. Applied Sciences. 2019; 9: 4500.
  • [32] Akay, B., Karaboga, D., Akay, R., A comprehensive survey on optimizing deep learning models by metaheuristics. Artificial Intelligence Review. 2022; 55: 829-894.
  • [33] Zhang, X., Shi, J., Zhu, G., Xiao, Y., Chen, H., Study of regional monthly precipitation based on CEEMD-BILSTM coupled model. Water Supply. 2022; 22: 8036-8049.
  • [34] Dey, R., Salem, F. M., Gate-variants of gated recurrent unit (GRU) neural networks. IEEE 60th International Midwest Symposium on Circuits and Systems. 2017; 1597-1600.
  • [35] Veloso, G. A., Loschi, R. H., Dynamic linear degradation model: Dealing with heterogeneity in degradation paths. Reliability Engineering & System Safety. 2021; 210: 107446.
  • [36] Frederick, D. K., DeCastro, J. A., Litt, J. S. (2007). User's Guide for the Commercial Modular Aero-Propulsion System Simulation (C MAPSS). NASA/ARL, Technical Manual TM2007-215026.
  • [37] Aydın, H., Effects of air to fuel ratio on parameters of combustor used for gas turbine engines: Applications of turbojet, turbofan, turboprop and turboshaft. Energy. 2024; 305: 132346.
  • [38] Babu, G. S., Zhao, P., Li, X. L., Deep convolutional neural network based regression approach for estimation of remaining useful life. International Conference on Database Systems for Advanced Applications. Springer, Cham. 2016; 214-228.
  • [39] Zheng, S., Ristovski, K., Farahat, A., Gupta, C., Long short-term memory network for remaining useful life estimation. 2017 IEEE International Conference on Prognostics and Health Management (ICPHM). 2017.
  • [40] Zhang, J., Wang, P., Yan, R., Gao, R. X., Long short-term memory for machine remaining life prediction. Journal of Manufacturing Systems. 2018; 48: 78-86.
  • [41] Azyus, A. F., Wijaya, S. K., Naved, M., Prediction of remaining useful life using the CNN-GRU network: A study on maintenance management. Software Impacts. 2023; 17: 100535
Toplam 41 adet kaynakça vardır.

Ayrıntılar

Birincil Dil Türkçe
Konular Bilgi Sistemleri Geliştirme Metodolojileri ve Uygulamaları, Mekatronik Sistemlerin Simülasyonu, Modellenmesi ve Programlanması
Bölüm Tasarım ve Teknoloji
Yazarlar

Mustafa Ozan Özkök 0009-0009-8098-1155

Rüştü Akay 0000-0002-3585-3332

Erken Görünüm Tarihi 29 Temmuz 2025
Yayımlanma Tarihi
Gönderilme Tarihi 9 Aralık 2024
Kabul Tarihi 10 Nisan 2025
Yayımlandığı Sayı Yıl 2025 Cilt: 13 Sayı: 3

Kaynak Göster

APA Özkök, M. O., & Akay, R. (2025). Derin Öğrenme Modelleri Kullanarak Endüstriyel Makinelerin Kalan Kullanım Ömrü Tahmini. Gazi Üniversitesi Fen Bilimleri Dergisi Part C: Tasarım Ve Teknoloji, 13(3). https://doi.org/10.29109/gujsc.1598534

                                     16168      16167     16166     21432        logo.png   


    e-ISSN:2147-9526