Sensitivity analysis for the optimal minimal repair/replacement policies under the framework of markov decision process

For the maintenance and replacement of system subject to stochastic deterioration, flexibility is an important consideration factor. Different possible maintenance policies are discussed in this paper where the failure distribution modeled by the Weibull distribution. The total discounted cost concept is very important in decision making for the complex system such as expensive capital equipment with operating life cycle lasting for years. A modified maintenance model is proposed with incorporating the costs of operating cost, the maintenance cost of minimal repairs, failure replacements and preventive replacements. We investigate the maintenance policies by modeling the deteriorating process of a system as a Markov decision process. Under the discounted cost criterion, the optimal parameters for the control limit policy are obtained. The sensitivity analysis of these control limits is performed for the flexibility of decision making. The optimal equations of the Markov decision process are solved by using the backward recursive scheme over a set of finite planning horizons to approximate the optimal policies for the infinite planning horizons.