Competing risk model for long-stop-short-run systems

Long-stop-short-run(LSSR) system is a kind of system that is longtime non-working and short time operating. It is important to study the reliability of LSSR system in some safety critical areas. There exist manifold failure mechanisms in the LSSR system. In this paper, a competing risk model is proposed to describe the reliability of the LSSR system. LSSR system presents two main kinds of failure modes: degradation failure and sudden failure. The degradation is described by stochastic Gamma process, and the sudden failure is described by Weibull distribution. In order to consider the influence of unknown degradation mechanisms to the reliability, Bayesian method is used to update the Weibull model parameters based on records of history values. The posterior distribution is used to estimate the expected reliability for the sudden failure mode. This competing risk model is illustrated by a pump case. Compared with the traditional method, the result of competing risk model is more appropriate for the reality.