An Algorithm for Simultaneous Optimization of Parameters of Condition-based Preventive Maintenance

Abstract

A new condition-based preventive maintenance model for a system, subject to deterioration-failures and to random failures with increasing intensity, is presented. Deterioration is modeled as discrete stages. After an inspection, based on the degree of deterioration, a minimal maintenance or a major maintenance is performed, or no action is taken. Deterioration failures are restored by complete replacements; Random failures are restored by minimal repair. Major maintenance restores the system to `τ' deterioration stages younger (τ > 1), while minimal maintenance restores the system one stage. The proposed model considers an accumulated deterioration based increasing intensity for the random failures. A continuously increasing failure rate (for example Weibull) is converted into a stepwise increasing failure rate using stair-step approximation. An exact recursive algorithm computes the steady-state probabilities of the system. Based on maximization of the system availability or minimization of total cost, an optimal inspection policy within this strategy, optimal inter-inspection time, and threshold deterioration level for replacement, are obtained. The effects of threshold deterioration level for deterioration replacements, inspection, maintenance, and repair parameters are investigated.