Degradation modeling-based remaining useful life estimation: A review on approaches for systems with heterogeneity

Prognostics and health management has drawn increasing attention and gained deepening recognition and widening applications during the past decades. Due to offering guidance for sequential managements involving inspection schedule, maintenance, replacement, and spare parts ordering, remaining useful life estimation has been termed as the kernel technology of prognostics and health management and is the focus of this research in the field of reliability. Heterogeneity is widespread in the inner states of a system and its related working environments. This article provides a review on approaches for degradation modeling and remaining useful life estimation, with an emphasis on the heterogeneity in the systems. Approaches for three kinds of heterogeneity, including the unit-to-unit variability, the variability in time-varying operating conditions, and the diversity of tasks and workloads of a system during its lifetime, are summarized consecutively, and the corresponding methods are provided. Merits and drawbacks are summed up, respectively, following each approach. In addition, several possible future research directions are provided at the end of this article.