Maintenance optimization has been extensively studied in the past decades. However, most of the existing maintenance models focus on single-component systems. Multi-component maintenance optimization, which joins the stochastic failure processes with the combinatorial maintenance grouping problems, remains as an open issue. To address this challenge, we study this problem in a finite planning horizon by i) developing a set of novel modeling techniques and building a two-stage stochastic integer model, and ii) based on its structural properties, designing and implementing an efficient heuristic algorithm under the progressive hedging framework. Comparing to three popular methods for stochastic integer programming, our algorithm demonstrates a drastically improved capacity in handling practically large-size problems. By using a rolling horizon scheme, our approach is further benchmarked with a conventional dynamic programming approach adopted in the literature. Numerical results show that the stochastic maintenance model and the designed heuristic can lead to significant cost savings.
Zhicheng Zhu, Yisha Xiang, Bo Zeng. Multi-component maintenance optimization: A stochastic programming approach. INFORMS Journal on Computing. (Under review).
View Multi-component Maintenance Optimization: A Stochastic Programming Approach