Dynamic imperfect condition-based maintenance for systems subject to nonlinear degradation paths
X. Zhao, Z. Liang, A. Parlikad and M. Xie
https://doi.org/10.19124/ima.2018.001.32
Abstract
This paper presents a maintenance optimization framework for systems suffering from nonlinear continuous degradation. The inspection interval is dynamically determined by the historical system conditions. We model the degradation path as a nonlinear Wiener process with time-varying drift parameter. Techniques to predict remaining useful life (RUL) is utilized to optimize maintenance policy by minimizing the expected cost rate. The effect of imperfect maintenance is assumed to be random in the sense that the maintenance action reduces the systems degradation level by a random proportion described by a beta distribution. Two thresholds on the degradation are determined for the preventive imperfect maintenance and perfect replacement, respectively. We evaluate expected cost rate using Monte Carlo simulation. A dataset from the real-world example is used to provide the pilot parameters as input for the optimization maintenance policies. Afterward, numerical examples are presented to illustrate the proposed method.