RUL Prediction of Bearings Based on Mixture of Gaussians Bayesian Belief Network and Support Vector Data Description
This paper presents a method to predict the Remaining Useful Life (RUL) of bearings based on theories of Mixture of Gaussians Bayesian Belief Network (MoG-BBN) and Support Vector Data Description (SVDD). In this method, the feature vectors, which are used to train the corresponding MoG-BBN and SVDD model, are extracted from raw sensor data by using wavelet packet decomposition (WPD). Genetic algorithm is employed to determine the initial value of the variables in MoG-BBN training algorithm so that the stability of MoG-BBN can be enhanced. The two models are combined to acquire a good generalization ability. We demonstrate the effectiveness of the proposed method by using actual bearing datasets from the NASA prognostic data repository.
KeywordsWavelet packet decomposition Mixture of gaussians bayesian belief network Genetic algorithm Support vector data description Remaining useful life
This work was partially supported by Chinese National Hi-Tech. R&D (863) Program under grant 2015AA042102.
- 1.Jammu, N.S., Kankar, P.K.: A review on prognosis of rolling element bearings. Int. J. Eng. Sci. Technol. 3(10), 7497–7503 (2011)Google Scholar
- 2.Lee, J., Qiu, H., Yu, G., Lin, J.: Rexnord Technical Services, IMS, University of Cincinnati. Bearing Data Set, NASA Ames Prognostics Data Repository. NASA Ames Research Center, Moffett Field, CA (2007). http://ti.arc.nasa.gov/project/prognostic-data-repository
- 6.Tobon-Mejia, D.A., Medjaher, K., Zerhouni, N., et al.: Hidden Markov models for failure diagnostic and prognostic. In: Prognostics and System Health Management Conference (PHM-Shenzhen), 2011, pp. 1–8. IEEE (2011)Google Scholar
- 10.Sloukia, F., El Aroussi, M., Medromi, H., et al.: Bearings prognostic using mixture of gaussians hidden markov model and support vector machine. In: 2013 ACS International Conference on Computer Systems and Applications (AICCSA), pp. 1–4. IEEE (2013)Google Scholar
- 12.Wald, R., Khoshgoftaar, T.M., Sloan, J.C.: Using feature selection to determine optimal depth for wavelet packet decomposition of vibration signals for ocean system reliability. In: 2011 IEEE 13th International Symposium on High-Assurance Systems Engineering (HASE), pp. 236–243. IEEE (2011)Google Scholar
- 13.Zhang, T., Zhang, H., Wang, Z.: Float encoding genetic algorithm and its application. J. Harbin Inst. Technol. 32(4), 59–61 (2000)Google Scholar