Health Maintenance Decisions Based on Hazard Rate Function under Degradation Process

Abstract

The classical Hazard Rate Function (HRF) is typically used to make preventative maintenance (PM) decisions and always estimates HRF based on time to failure data. Even so, PM determination using conventional HRF is not suitable for systems with condition monitoring. The system is stopped checking and maintaining when the measured health state exceeds a predetermined threshold, which is always referred to as a soft failure. Since the application of HRF for PM decision making has great advantages, the classic HRF should be extended to combine soft and hard failure of systems with condition monitoring. In this paper, we define the HRFs of hard and soft failure for system under condition monitoring and propose a method to estimate the HRFs with data of failure time. We discuss in detail the relationship between these two HRFs and the classical HRF. With double stochastic processes (processes of degradation and measured healthy status), the properties of these HRFs are also researched. Further the optimal maintenance decisions are made for non-repairable and repairable systems upon these two types of HRFs. Eventually, the idea of this paper is verified by numerical examples.