Abstract
This paper considers condition-based maintenance and spare parts inventory policy simultaneously for a system consisting of different machines and components. The degradation of the components are modeled by Gamma process. By the sensors on the components, we continuously monitor the degradation level of the components, and once the degradation level exceeds the predefined degradation thresholds, imperfect repair maintenance or replacement maintenance are performed. A simulation-based optimization approach is proposed to find the optimal components inventory policy and degradation level thresholds of components. The proposed approach is based on Stochastic Trust-Region Response Surface Method (STRONG), coupled with the Kriging metamodel and the Nelder-Mead simplex method. A numerical study shows that the proposed model and the method can achieve minimized maintenance cost as expected.