Beijing

Covariate modeling including system functioning environment and system operating factors is regarded as an effective approach to explore the underlying mechanism of a reparable system. The covariate information is used for monitoring state changes to the functioning of the system so that the system performance (or capacity) can be judged and decisions on maintenance actions can adjusted accordingly. Cox (1972) initiated PH model, which quickly generated medical applications during 1980s but the industrial applications of the PH model were developed somewhat later, say, Dale (1985), Love and Guo (1991), Ascher and Kobbacy (1998) etc. Andersen, Borgan, Gill, and Keiding (1993) provided a solid theoretical foundation and an applicable methodology for covariate modeling based on counting processes. Guo and Love (2003) proposed a fuzzy covariate model for repairable system which is based on fuzzy logical function and static in nature. In this paper, we will investigate the system behavior under covariate influence in terms of grey system theory, specifically, the static GM(0,h) and dynamic GM(1,h) models. An industrial data is used for illustrative purpose