The Joint Research On Delay Time Based System Preventive Maintenance Modeling

In production enterprises,preventive maintenance(PM)is often needed to avoid the sudden failure of some key equipment and components,and to ensure continuous production.So far,although many achievements have been made in the research on preventive maintenance of equipment both in theory and applications,they have lots of shortcomings.Based on the delay time theory,this paper studies the preventive maintenance of equipment more deeply from the following four aspects.The main research contents and innovation points are as follows:(1)A joint decision preventive maintenance model for age replacement and fault replacement.The system's state is divided into two stages:normal stage and defective state.It is assumed that the system is affected by two kinds of failure modes,one is random failure,and the other is delay time failure.The system is replaced in three cases:defect replacement,preventive replacement and failure replacement.The optimal strategy is obtained by calculating the expected cost and expected update length of various possible cases,and a case analysis is carried out to verify the validity of the model.Furthermore,on the basis of the fixed diagnosis period,the adjustable diagnostic strategy is considered and compared with the original model.The result shows that the adjustable strategy is better than the fixed one.(2)The influence of failures' correlation on optimal strategy is considered.It is assumed that the system studied is a single component system with two fault modes:one is the traditional 0-1 failure,the other is the delay time failure mode.However,the two fault modes are related,that is,when the system enters the defect state in the delay time failure mode,it will affect the failure rate in the 0-1 fault mode.The joint failure rate of the two failure modes is represented by the copula joint distribution function.By calculating the expected cost and expected update length of all possible updates,we get the best preventive maintenance policy.(3)An advanced maintenance strategy based on delay time theory is considered.The system considered is consisting of two components,and component 1 is the key component.When the component 1 fails,the whole system is updated.Meanwhile,in order to prevent the failure of component 1,preventive maintenance activities need to be taken.Compared with component 1,component 2 is not the key component,which is cheaper and easier to replace.When componnet 2 fails,only this component itself is replaced without affecting the update of the system.However,if its failure occurs within a given system preventive replacement cycle threshold,the system is updated in advance to avoid overreplacement of componnet 2.Therefore,a joint decision framework is established to prevent preventive maintenance cycle and update threshold in advance.(4)A preventive maintenance model for the repairable system was established as the supplementary research for the preventive maintenance modeling.Consider a single-unit repairable system.Assume that the component's state variables can be monitored.When the component's state exceeds a certain threshold,the componnet will be repaired imperfectly,ie,between repairs and repairs.Due to imperfect maintenance,the degradation of components will gradually increase.In order to prevent over-maintenance,it is assumed that when the number of maintenance reaches a certain amount,the system will be replaced.This process can be described by a non-homogeneous Poisson process.In addition,based on the life information provided by the manufacturer,in order to prevent system failure,the system will be replaced based on the age of service.This model takes the actual degradation process of a fan system in a steel plant as an example.According to the actual maintenance records and replacement cycle,the parameters of the model are estimated and hypothesized.Finally,the effectiveness of the model is verified through optimization.