Prediksi Failure Degradation Bearing dengan Pendekatan Gabungan Model
Keywords:
Bearing, Cox Proportional Hazard (CPH), Support Vector Regression (SVR)
Abstract
Abstract. The reliability of machine components is a very important aspect to consider, one of the critical components in a machine is the bearing, which serves as a support and allows rotating movement between two components. Failure in bearings can result in serious damage to the machine and stop operations, this has a major impact on productivity and maintenance costs. Predictions were made using the Support Vector Regression method with the Cox Proportional Hazard model as the target vector. The Cox Proportional Hazard model is used to see the effect of covariates on bearing survival time, while Support Vector Regression with RBF kernel is used to predict survival chances based on the risk from the Cox model. In this study, the Weibull distribution hazard function is used as the baseline hazard function in the Cox model, to calculate the value of the shape and scale parameters in the Weibull distribution, the Least Squares method is used until the respective values of the shape parameter and scale parameter are 1.02045 and 7.3732. The results show that combining the Cox Proportional Hazard model and Support Vector Regression can provide a high average accuracy value of 99.83% on training data and 98.82% on testing data. As in the journal article (Caesarendra et al., 2010) with a high accuracy value, it can be said that the model can be used. Abstrak. Keandalan komponen mesin menjadi aspek yang sangat penting untuk diperhatikan, salah satu komponen kritis dalam mesin adalah bearing, yang berfungsi sebagai penopang dan memungkinkan pergerakan berputar antara dua komponen. Kegagalan dalam bearing dapat mengakibatkan kerusakan serius pada mesin dan menghentikan operasi, hal ini berdampak besar pada produktivitas dan biaya pemeliharaan. Dilakukan prediksi menggunakan metode Support Vector Regression dengan model Cox Proportional Hazard sebagai target vektornya. Model Cox Proportional Hazard digunakan untuk melihat pengaruh dari kovariat terhadap waktu survival bearing, sedang Support Vector Regression dengan kernel RBF digunakan untuk memprediksi peluang survival berdasarkan risiko dari model Cox. Dalam penelitian ini digunakan fungsi hazard berdistribusi Weibull sebagai fungsi baseline hazard dalam model Cox, untuk menghitung nilai parameter bentuk dan skala pada distribusi Weibull digunakan metode Kuadrat Terkecil, hingga diperoleh nilai masing-masing dari parameter bentuk dan parameter skala adalah 1,02045 dan 7,3732. Hasil menunjukkan bahwa kombinasi model Cox Proportional Hazard dan Support Vector Regression dapat memberikan nilai rata-rata akurasi yang tinggi sebesar 99,83% pada data training dan 98,82% pada data testing. Seperti pada artikel jurnal (Caesarendra et al., 2010) dengan nilai akurasi yang tinggi, maka dapat dikatakan bahwa model dapat digunakan.References
Awad, M., & Khanna, R. (2015). Efficient Learning Machines: Theories, Concepts, and Applications for Engineers and System Designers. Apress.
Caesarendra, W., Widodo, A., & Yang, B. S. (2011). Combination of probability approach and support vector machine towards machine health prognostics. Probabilistic Engineering Mechanics, 26(2), 165-173. https://doi.org/10.1016/j.probengmech.2010.09.008.
Cox, D.R., Oakes, D., & Oakes, D. (1984). Analysis of survival data. Taylor & Francis.
Ebeling, C. E. (1997). An Introduction to Reliability and Maintainability Engineering. McGraw Hill.
Hedianti, E. S. (2019). Peramalan Harga Saham Dengan Menggunakan Metode Support Vector Regression (SVR).
Khotimah, C. (2018). Additive Survival Least Square SVM Untuk Analisis Data Survival (Studi Simulasi dan Studi Kasus pada Data Pasien Kanker Serviks di RSUD dr. Soetomo Surabaya).
Kleinbaum, D. G., & Klein, M. (2012). Survival Analysis: A Self-Learning Text, Third Edition. Springer New York.
Nikulin, M., & Wu, H.-D. I. (2016). The Cox Model and Its Applications. Springer Berlin Heidelberg. https://link.springer.com/book/10.1007/978-3-662-49332-8.
Otaya, L. G. (2016). Distribusi Probabilitas Weibull Dan Aplikasinya (Pada Persoalan Keandalan (Reliability) dan Analisis Rawatan (Maintainability). Tadbir: Jurnal Manajemen Pendidikan Islam, 4(2), 44-66. https://journal.iaingorontalo.ac.id/index.php/tjmpi/article/view/438.
Razali, A. M., Salih, A. A., & Mahdi, A. A. (2009). Estimation Accuracy of Weibull Distribution Parameters. The Journal of Applied Sciences Research, 5, 790-795.
Caesarendra, W., Widodo, A., & Yang, B. S. (2011). Combination of probability approach and support vector machine towards machine health prognostics. Probabilistic Engineering Mechanics, 26(2), 165-173. https://doi.org/10.1016/j.probengmech.2010.09.008.
Cox, D.R., Oakes, D., & Oakes, D. (1984). Analysis of survival data. Taylor & Francis.
Ebeling, C. E. (1997). An Introduction to Reliability and Maintainability Engineering. McGraw Hill.
Hedianti, E. S. (2019). Peramalan Harga Saham Dengan Menggunakan Metode Support Vector Regression (SVR).
Khotimah, C. (2018). Additive Survival Least Square SVM Untuk Analisis Data Survival (Studi Simulasi dan Studi Kasus pada Data Pasien Kanker Serviks di RSUD dr. Soetomo Surabaya).
Kleinbaum, D. G., & Klein, M. (2012). Survival Analysis: A Self-Learning Text, Third Edition. Springer New York.
Nikulin, M., & Wu, H.-D. I. (2016). The Cox Model and Its Applications. Springer Berlin Heidelberg. https://link.springer.com/book/10.1007/978-3-662-49332-8.
Otaya, L. G. (2016). Distribusi Probabilitas Weibull Dan Aplikasinya (Pada Persoalan Keandalan (Reliability) dan Analisis Rawatan (Maintainability). Tadbir: Jurnal Manajemen Pendidikan Islam, 4(2), 44-66. https://journal.iaingorontalo.ac.id/index.php/tjmpi/article/view/438.
Razali, A. M., Salih, A. A., & Mahdi, A. A. (2009). Estimation Accuracy of Weibull Distribution Parameters. The Journal of Applied Sciences Research, 5, 790-795.
Published
2025-02-03
Section
Articles