Research On Condition-based Maintenance Policy Optimization Under Imperfect Maintenance

Maintenance decision-making has emerged as an important area of industrialresearch. Recently, condition-based maintenance is becoming popular and highlightedby researchers and practitioners, since it can help improving maintenance efficiency andavoiding excessive or insufficient maintenance. Most of the existing maintenancemodels allow finding the optimal maintenance policy; however, few of them areproposed to consider the imperfectness of maintenance actions during the optimizationof maintenance policy. The objective of this work is to propose a new structure ofcondition-based maintenance policy model which consider the imperfectness ofmaintenance actions.This paper considers the condition-based maintenance policy model both for singleunit and two-unit deteriorating system. Each unit is subject to gradual deterioration andis monitored by non-periodic inspections. For a system which gradually deteriorates, itis natural in an attempt to avoid failure and its possible negative effects to consider thesystem condition in order to carry out preventive maintenance actions.A new condition-based and opportunistic maintenance policy is proposed in thispaper. Under this maintenance policy, the deterioration processes of equipment arediscrete and modeled using continuous-time Markov chains. And the maintenancedecisions are given based on deterioration states. The maintenance decisions includemaintenance actions and inspection intervals for different states of equipment. Theassociated cost models are developed help evaluating the long-run average cost basedon the stochastic properties of the deterioration process of the maintained system.Different maintenance policies can be compared using the cost model to find theoptimal policy with minimal cost.By numerical examples, the application of the proposed maintenance policy to theoptimization of condition-based maintenance is illustrated.