Research On Reliability Modeling And Maintenance Decision Of Repairable System

With the rapid development of industrial field,the structure of the system is getting more and more complex and precise,showing the characteristics of dependence and polymorphism.The situation makes the maintenance of system face more and more challenges.The traditional maintenance method is based on fault repair or regular preventive maintenance.It has two shortcomings,on one hand,insufficient maintenance efforts,resulting in system overload operation,system reliability is reduced,the level of failure increased,thus requiring more maintenance costs.On the other hand,the overmaintenance resulting from short-cycle preventive maintenance or preventive replacements of the system result,which also requires more costs.Therefore,the traditional maintenance method can't meet the maintenance requirements of complex system.In this thesis,for the problems existing in the traditional maintenance methods,we combine the system maintenance and decision optimization,and we carry out the following three aspects of research.(1)For the difference of component opportunistic maintenance time threshold and the high cost of opportunistic maintenance,the opportunistic maintenance model based on the imperfect maintenance is proposed.Firstly,based on the minimum maintenance time threshold,imperfect maintenance time threshold and replacement maintenance time threshold,the opportunistic maintenance model is established.Then we establish a decision model based on imperfect maintenance.By means of heuristic algorithm,we find out the imperfect maintenance time threshold and replacement maintenance time threshold of the component.And we can get the maintenance strategy.By establishing an opportunistic maintenance model based on imperfect maintenance for an aeroengine system,the maintenance cost per unit time of the system is reduced.(2)For the multi-tasking of system and the uncertainty of working time,the selective maintenance model with multi-task uncertainty is proposed.We divide the multi-tasking time period by dynamic programming.So the selective maintenance decision problem of multi-task uncertainty can be transformed into the certain selective maintenance decision problem.We can get the optimal time slice division and maintenance strategy by traversing the results of the time period division.Through a multi-task uncertainty selective maintenance model for a coal transportation system,the maintenance cost is reduced and the effectiveness of the proposed method is verified.(3)For the problem that single-objective decision optimization leads to the extreme maintenance strategy,a bi-objective selective maintenance model of multi-state system is proposed.Firstly,the degradation model of multi-state components is established.Secondly,the decision model of minimum maintenance cost and maximum system performance are established respectively.Then the two single-objective decision models are transformed into one bi-objective decision model.Finally,the Pareto optimal set of maintenance strategy is found according to the bi-objective decision model.The experimental results show that the multi-state system selective maintenance model based on bi-objective decision can find a compromise between maintenance cost and task completion rate,which verifies the effectiveness of the model.