Optimization of a Failure – prone Manufacturing System with Regular Preventive Maintenance: an IPA Approach

Abstract

AbstractThis paper addresses the optimization of a single-stage single-product manufacturing system with a constant demand and a systematic preventive maintenance policy. The manufacturing system is modeled by a continuous-flow model. The machine is subject to time-dependent failures and its production speed is given by a hedging point policy. Times to failure and times to repair are random variables with exponential distribution. We propose a simulation-based optimization method for determining the optimal buffer level in order to minimize the long run average cost including inventory holding cost, backordering cost and corrective and preventive maintenance costs. The optimization algorithm is based on the Infinitesimal Perturbation Analysis (IPA) technique for estimation of gradients along the simulation. The unbiasedness of the IPA estimators is established. Numerical results based on the simulation algorithm are proposed to determine the optimal buffer level and to derive the optimal value of the period of the systematic preventive maintenance.