STUDY OF FIXED EFFECT AND RANDOM EFFECT GAMMA DEGRADATION MODELS
https://doi.org/10.21285/1814-3520-2018-6-120-128
Abstract
PURPOSE. Construction of the random effect gamma degradation model results in the enlarged dimension of the vector of estimated parameters. Therefore, it is necessary to understand whether the introduction of the random effect allows to obtain more accurate estimates of model parameters or not. The paper compares the fixed effect and random effect gamma degradation models in terms of the estimation accuracy of their parameters. METHODS. The estimation accuracy of the highest likelihood of degradation model parameters is studied through the calculation of the Euclidean norm of the relative error of the obtained estimates for the different values of random effect based on the Monte-Carlo method. RESULTS AND THEIR DISCUSSION. The estimation accuracy of parameters is higher for the random effect degradation model with the increase of the random effect value. However, if unit-to-unit variability coefficient does not exceed 0.1, it is recommended to use the fixed effect degradation model for the higher estimation accuracy of model parameters. MAIN CONCLUSIONS. It is shown that the use of the random effect gamma degradation model is not advisable in the case of small sample volumes and an insignificant random effect.
References
1. Антонов А.В., Никулин М.С. Статистические модели в теории надежности // М.: Абрис, 2012. 390 с.
2. Tsai C.-C., Tseng S.-T., Balakrishnan N. Mis-specification analyses of gamma and Wiener degradation processes // Journal of Statistical Planning and Inference. 2011. No. 12. P. 25-35.
3. Chimitova E.V., Chetvertakova E.S., Sergeeva S.A., Osinceva E.A comparative analysis of the wiener, gamma and inverse gaussian degradation models // Applied methods of statistical analysis. Nonparametric methods in cybernetics and system analysis (AMSA'2017): proceedings of the International Workshop (Krasnoyarsk, 18-22 September 2017). Krasnoyarsk, 2017. P. 160-167.
4. Wang X., Balakrishnan N., Guo B. Mis-specification Analyses of Nonlinear Wiener Process-based Degradation Models, Communications in Statistics - Simulation and Computation. 2016, 45. P. 814-832.
5. Zhang C., Lu X., Tan Y., Wang Y. Reliability demonstration methodology for products with Gamma Process by optimal accelerated degradation testing // Reliability Engineering and System Safety. 2015. Vol. 142. P. 369-377.
6. Tsai T.R., Sung W.Y., Lio Y.L., Chang S.I., Lu J.C. Optimal two-variable accelerated degradation test plan for gamma degradation processes // IEEE Transactions on Reliability. 2016. Vol. 65. No. 1. P. 459-468.
7. Чимитова Е.В., Четвертакова Е.С. Построение гамма деградационной модели надежности с учетом влияния объясняющих переменных // Вестник Томского государственного университета. Управление, вычислительная техника и информатика. 2014. № 4(29). C. 51-60.
8. Wang X., Jiang P., Guo B., Cheng Z. Real-time Reliability Evaluation for an Individual Product Based on Change-point Gamma and Wiener Process // Quality and Reliability Engineering International. 2014. P. 513-525.
9. Chimitova E.V., Chetvertakova E.S. Statistical degradation models for reliability analysis in non-destructive testing // IOP Conference Series: Materials Science and Engineering. 2017. Vol. 189. 6 p.
10. Ye Zh.-Sh., Xie M., Tang L.-Ch., Chen N. Semiparametric Estimation of Gamma Processes for Deteriorating Products // Technometrics. 2016. Vol. 56. No. 4. P. 504-513.
11. Tang S., Guo X., Yu C., Xue H., Zhou Z. Accelerated degradation tests modeling based on the nonlinear wiener process with random effects // Mathematical Problems in Engineering. 2014. 11 p.
12. Chimitova E.V., Chetvertakova E.S. A comparison of the “fixed-effect” and “random-effect” gamma degradation models // Applied methods of statistical analysis. Nonparametric approach, AMSA’2015: рroceedings of the international workshop (Novosibirsk 14-19 September 2015). Novosibirsk, 2015. P. 161-169.
13. Tsai C.C., Tseng S.T., Balakrishnan N. Optimal Design for Degradation Tests Based on Gamma Processes with Random Effects // IEEE Trans. Reliab. 2012. Vol. 61. P. 604-613.
14. Микер У.К., Доганаксой Н., Хан Дж. Дж. Использование данных о деградации для анализа надежности изделий // Методы менеджмента качества. 2009. № 4.
15. Nikulin M., Bagdonavicius V. Accelerated Life Models: Modeling and Statistical Analisys // Boca Raton: Chapman & Hall/CRC. 2001. 334 p.
Review
For citations:
Chetvertakova E.S. STUDY OF FIXED EFFECT AND RANDOM EFFECT GAMMA DEGRADATION MODELS. Proceedings of Irkutsk State Technical University. 2018;22(6):120-128. (In Russ.) https://doi.org/10.21285/1814-3520-2018-6-120-128
Views:
171
Email this article 