Joint Optimization Policy Of Condition-Based Preventive Maintenance And Spare Parts Management For Equipment

The demand for spare parts is driven by maintenance requirements and maintenance activities depend on the availability of spare parts,and spare parts affect the development of maintenance activities.In particular,for a critical or expensive industrial plant it is common to at most order and store one spare,rather than lot size.So far,there have been relatively few studies on the joint optimization of preventive maintenance and spare parts ordering for equipment and component,and there are still some shortcomings in these studies in theory and applications.Based on the renewal process theory,the joint optimization of maintenance and spare parts ordering is studied in the thesis with a single component(critical component).The major research works are as follows:(1)This part proposes a joint optimization strategy of condition-based maintenance and spare parts ordering based on non-periodic inspection.The two-stage delay time theory(normal stage,defect stage)is used to describe the degradation process,based on the aperiodic state detection strategy(T1,T2),considering the imperfect maintenance of the defect state and the spare part order of the random lead time(normal order,emergency order)decision-making,propose a joint optimization strategy for condition-based maintenance and spare parts ordering of single-component(key component)system;based on the maintainability of components,use the proportional age reduction model for imperfect maintenance;Inspection cost,imperfect maintenance cost,cost of system failure,emergency ordering,regulation ordering,holding cost and loss of shortage are taken been account and a long-term expected cost rate optimization model is established to optimate the optimal inspection interval(T1,T2)and the maximum number of IPM Nmax.Alternatively,a comparative model in which the spare is ordered at time 0 is also constructed.The proposed policy is illustrated by the numerical example and the results indicate that it decreases the expected cost per unit time compared to the comparative model.(2)This part proposes a joint optimization strategy of condition-based maintenance and spare parts ordering based on multiple preventive maintenance.Based on the three-stage fault process(normal phase,light defect phase,severe defect phase),by considering the delay of maintenance activities caused by spare parts delay,a joint optimization strategy of conditionbased maintenance and spare parts ordering is proposed,where imperfect maintenance(IPM)and preventive maintenance(IPM reaches the threshold renewal,ageing renewal)and fault repair,spare parts regular order and expedited ordering are considered simultaneously.Inspection,maintenance and spare parts related costs are been account,and the long-term expected cost rate optimization model is established.The Monte Carlo-PSO hybrid algorithm and the enumeration algorithm are designed to simulate and optimize the model respectively.The optimal inspection interval,age-replacement threshold and maximum imperfect maintenance threshold are obtained when the long-term expected cost rate is the lowest.Finally,a numerical example is given to illustrate the proposed joint optimization policy and the sensitivity analysis of the parameters is carried out.The Monte Carlo-based PSO solution algorithm was evaluated to be faster than the enumerated optimization algorithm in solving the model,and the effect of the imperfect maintenance degree on the results was studied.(3)This part proposes a joint policy of condition-based maintenance and spare ordering by predicting the remaining useful life(RUL).Based on the Wiener process,considering the three control limits of preventive replacement threshold,fault threshold and ordering threshold,and the optimization model of the proposed joint policy is established to minimize the longterm expected cost rate.A discrete-event simulation algorithm is designed to optimate the optimal ordering threshold and preventive replacement threshold,and the parameter sensitivity analysis is carried out.Finally,an illustrate example is applied to demonstrate the applicability and effectiveness of the proposed model and algorithm.