Reliability Analysis Using Real Options Based on Hierarchical Bayesian Methods

抄録

Real option approaches are handled for reliability evaluations based on degradation models. In general, the reliability analysis is focused on a random variable of life data, and the Weibull distribution is used as a reliability modeling. However a degradation process is significant for making value of the reliability. In particular, stochastic process models are useful for representations of the degradation paths. In this paper, a hierarchical Bayesian VG (Variances Gamma) model is proposed for the representation of the degradation phenomenon, and the real option method based on a simulation is applied to the reliability evaluation using the model. It is also shown that other reliability approaches for the degradation analysis are handles based on GARCH (Generalized Autoregressive Conditional Heteroskedasticity) and SV(Stochastic Volatility) models, and a method of comparison for the models based on the information criterion EIC is proposed.