The Finite-time-domain Optimal Periodical Preventive Maintenance Policy With Warranty

As the users of appliances take more and more account on thewarranty as the added value of the appliances, and raise more and moreclaim on warranty, more and wider researches on the preventivemaintenance policy based on warranty should be taken into consideration.Standing on the users’ perspective, this paper goes on two steps whendiscussing about the periodical preventive maintenance under differentconditions within and post warranty in the finite time domain.Firstly, the paper considers the problem of different failure rate withinand post warranty. And the idea of Upgrade Service is raisedhypothetically, which means providers of appliances will provide a largeand free preventive maintenance at the time point of the end of warranty sothat they can reject the demands of warranty extension. So the failure ratepost warranty should be smaller than that within warranty. Secondly, basedon the problem discussed before, the paper deeply considers the otherproblem of different period within and post warranty. This paper proposedthe optimal preventive maintenance policy based on the length of warrantyand period as the decision variables, and the minimization of the expectedtotal maintenance cost as the objective function. It is determined that theappliances can be freely repaired within warranty, while be paid by usersthemselves post warranty. However, the penalty cost made by the failureshould be paid by the users themselves both within warranty and postwarranty, as well as the cost of preventive maintenance. Therefore, therepair cost post warranty, the penalty cost and preventive maintenance cost within and post warranty are all included in the total maintenance cost.What’s more, the appliance can be minimally repaired when fail. The typeof the preventive maintenance is imperfect one, and the type of restorationis failure rate reduction. Finally, via the numerical examples, the feasibilityand validity of the models are verified successfully. In addition, the paperanalyzes the sensitivity of the basic parameters to see the influence on themodels.