A predictive maintenance system for hybrid degradation processes
International Journal of Quality & Reliability Management
ISSN: 0265-671X
Article publication date: 7 August 2017
Abstract
Purpose
The purpose of this paper is to propose a predictive maintenance (PdM) system for hybrid degradation processes with continuous degradation and sudden damage to improve maintenance effectiveness.
Design/methodology/approach
The PdM system updates the degradation model using partial condition monitoring information based on degradation type judgment. In addition, an extended multi-step-ahead updating stopping condition is adopted for performance enhancement of the PdM system.
Findings
An extensive numerical investigation compares the performance of the PdM system with the corresponding preventive maintenance (PM) policy. By carefully choosing the updating stopping condition, the PdM policy performs better than the corresponding PM policy.
Research limitations/implications
The proposed PdM system is applicable to single-unit systems. And the continuous degradation process should be well modeled by the stochastic linear degradation model (Gebraeel et al., 2009).
Originality/value
In literature, there are abundant studies on PdM policies for continuous degradation processes. However, research on hybrid degradation processes still focuses on condition-based maintenance policy and a PdM policy for a hybrid degradation process is still unreported. In this paper, a PdM system for hybrid degradation processes with continuous degradation and sudden damage is proposed. The PdM system decides PM schedules by fully utilizing the condition monitoring data of each specific product, and can hopefully improve maintenance effectiveness.
Keywords
Citation
You, M.-Y. (2017), "A predictive maintenance system for hybrid degradation processes", International Journal of Quality & Reliability Management, Vol. 34 No. 7, pp. 1123-1135. https://doi.org/10.1108/IJQRM-08-2016-0141
Publisher
:Emerald Publishing Limited
Copyright © 2017, Emerald Publishing Limited