Maintenance Optimization of Multi-Component Systems Subjected to <tex>$s$</tex>-Dependent Competing Risks and Imperfect Maintenance

Lin Zhang,Abdelhakim Khatab,Xiaohui Chen
DOI: https://doi.org/10.1109/IRASET57153.2023.10153017
2023-01-01
Abstract:This paper develops a novelreliability-based maintenance optimization approach in multi-componentsrepairable system (MCRS) subjected to both <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$s$</tex> -dependent competing risks andimperfect maintenance. The system operates a sequence of preventive maintenance(PM) cycles at the end of which a preventive replacement of the system iscarried out. Duration of each PM cycle is determined by a reliability thresholdwhich triggers either an imperfect PM or a preventive replacement; there are asmuch PM cycles as reliability thresholds. The objective of the proposedmaintenance approach is to determine the joint optimal reliability thresholdsand the number of PM cycles that minimize the total expected maintenance andbreakdown cost rate in the infinite time horizon. Conditions to derive optimalsolutions are formally established and discussed. Given the complexity of theresulting optimization problem, a fix-and-optimize numerical algorithm usingthe gradient descent method is developed. The numerical results obtained fromthe case study demonstrate the accuracy and the added value of the proposedapproach. These results clearly show that it is not only cost-effective butalso adaptive and allow then to increase the system availability. Indeed,except for the last PM cycle, it is shown that the optimal reliabilitythreshold increases by the increasing of the number of PM cycles.
What problem does this paper attempt to address?