Maintenance effect is a peculiar factor applied to repairable systems. Malik and Brown, Mahoney and Sivazlian (BMS) proposed general approaches to model maintenance effect, where each maintenance reduces the age of unit with respect to the rate of occurrences of failures. The main difference between Malik's and BMS's approaches is that each maintenance is assumed to reduce the elapsed time from the previous maintenance in Malik's model (Model Ⅰ), whereas it is assumed that maintenance reduces system age in BMS's model (Model Ⅱ). In this article, maintenance effect is represented by the age reduction function of explanatory variables such as maintenance time and cost. It describes situations that the more we invest on maintenance, the higher its effect is. The method of maximum likelihood method is used to estimate the age reduction function and the Weibull intensity function. Both Model Ⅰ and Ⅱ and two cases are considered: Case A - effective CM without. Case B - minimal CM and effective PM. Simulation results are presented to illustrate the accuracy and properties of estimation.
展开▼
