A PM policy optimization for a two-stage failure process with multiple modes of external shocks in random environment

This article proposes an adaptive sequential preventive maintenance (PM) policy for which an improvement factor is newly introduced to measure the PM effect at each PM. For this model, the PM actions are conducted at different time intervals so that an adaptive method needs to be utilized to determine the optimal PM times minimizing the expected cost rate per unit time. At each PM, the hazard rate is reduced by an amount affected by the improvement factor which depends on the number of PM's preceding the current one. We derive mathematical formulas to evaluate the expected cost rate per unit time by incorporating the PM cost, repair cost, and replacement cost. Assuming that the failure times follow a Weibull distribution, we propose an optimal sequential PM policy by minimizing the expected cost rate. Furthermore, we consider Bayesian aspects for the sequential PM policy to discuss its optimality. The effect of some parameters and the functional forms of improvement factor on the optimal PM policy is measured numerically by sensibility analysis and some numerical examples are presented for illustrative purposes.

Analytical method for optimization of maintenance policy based on available system failure data

HeA new machine will not fail easily in the early stage of its useful life. This phenomenon is consistent with the fact that the optimal sequential preventive maintenance (PM) policy which has longer PM intervals in its earlier stage of lifetime. In this paper, we propose a failure-rate-reduction periodic PM model with delayed initial time in a finite time span. Then, the optimal periodic PM policy is developed by minimizing the expected total maintenance cost, which can have smaller expected total maintenance cost than the optimal policy of the original failure-rate-reduction periodic PM model. The algorithm of finding the optimal PM policy for the proposed PM model is developed. Finally, examples are illustrated to verify the optimal policies of the proposed new PM model.

PurposeThis paper considers and discusses analytically the optimal preventive maintenance (PM) policies of aged plants such as fossil‐fired power plants.Design/methodology/approachShocks are assumed to occur at a nonhomogeneous Poisson process and the total damage due to each shock is additive. The system undergoes the PM at a certain time or the total damage exceeds a managerial level. The expected cost rate until PM is derived and optimal policies which minimizes it are discussed.FindingsThere exists a unique optimal time (T*) or managerial level (k*) which minimizes the expected cost rate. But there does not exist a positive pair (T*, k*), simultaneously.Research limitations/implicationsThe damage occurrence distribution is assumed to be nonhomogeneous Poisson one.Practical implicationsUseful methods to consider the optimal PM policies for power plant engineers.Originality/valueThis paper contributes to users of aged power plants economically and practically.

Preventive maintenance (PM) can slow the deterioration process of a repairable system and restore the system to a younger age or state. Many researchers focus on studying PM models for the cases of age or failure rate reduction and developing optimal PM policies. However, the PM actions, such as cleaning, adjustment, alignment, and lubrication work, may not always reduce system's age or failure rate. Instead, it may only reduce the degradation rate of the system to a certain level. Furthermore, most of the existing optimal PM policies result in very low reliability at the time of preventive replacement (PR). In practice, however, high reliability is usually required for a system to avoid failures being occurred. This paper is to develop an optimal periodic PM policy over an infinite time span by minimizing the expected cost rate with the consideration of reliability limit (R <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">min</sub> ,) for Weibull-failure-time systems with degradation rate reduction after each PM. The proposed optimal periodic PM policy consists of two PM models. One model has fully periodic PM time-interval for every preventive replacement cycle; the other model has partially periodic PM time-interval where the time interval between the final PM and the PR is not a constant. For specified reliability limit, the proper model is chosen by using the algorithm provided in this paper. Examples are demonstrated and the sensitivity analysis is also presented for the proposed PM models.

Traditional maintenance planning problems usually presume that preventive maintenance (PM) policies will be executed exactly as planned. In reality, however, maintainers often deviate from the intended PM policy, resulting in unpunctual PM executions that may reduce maintenance effectiveness. This article studies two imperfect PM policies with unpunctual executions for infinite and finite planning horizons, respectively. Under the former policy, imperfect PM actions are periodically performed and the system is preventively replaced at the last PM instant. The objective is to determine the optimal number of PM actions and associated PM interval so as to minimize the long-run average cost rate. However, the latter policy specifies that a system is subject to periodic PM activities within a finite planning horizon and there is no PM activity at the end of the horizon. The aim is then to identify the optimal number of PM activities to minimize the expected total maintenance cost. In this article, we discuss the modeling and optimization of the two unpunctual PM policies and then explore the impact of unpunctual executions on the optimal PM decisions and corresponding maintenance expenses in an analytical or numerical way. The resulting insights are helpful for practitioners to adjust their PM plans when unpunctual executions are anticipated.

Non-periodic preventive maintenance policies for complex multi-component products sold with loss-based two-dimensional warranties

Optimal preventive maintenance policy for leased equipment using failure rate reduction

In this paper, we deal with the problem of optimal schedules of two periodic imperfect preventive maintenance (PM) policies which use two types of PM models discussed in Doyen and Gaudoin (IEEE Trans Reliabil 35:71–75, 1986). The PM action of one PM model has an effect on the relative wear-out wear-out since the last PM action, while the PM action of another PM model is effective in restoring global wear-out since equipment started operating. For each PM policy, we derive formulas to compute the expected cost rate per unit time and discuss the optimal PM schedule which minimizes the expected cost rate per unit time. We also compare analytically the optimal schedules of two PM policies for various cost structures and investigate numerically the sensitivity of the cost structures and PM models to optimal schedules.

For a repairable and deteriorating system, the theoretical preventive maintenance (PM) models are usually complicated and need numerical methods to obtain the optimal PM policies since the system's hazard rate function is changed after each PM. It makes the application of the theoretical model not quite suitable for real cases. Moreover, the theoretical optimal PM solution is obtained by assuming the system is evaluated over an infinite time span. Yet, in reality, the deteriorating systems always have a finite life time. Hence, searching for an optimal PM solution in a finite time span might not be rational for a deteriorating system. Therefore, we consider using the simulation method to obtain a range of the near-optimal PM policy. The critical step of the simulation method for obtaining a near-optimal PM policy is the generation of the random variates (RV). Since the rejection method is a simple and easy-to-use method in simulating the failure process of a PM model, it is applied to generate the RVs of the time-between-failures (TBF) for the finite-time-span PM model with age reduction effect. In this research, examples of the near-optimal PM policies are presented and discussed.

The paper proposes equipment maintenance strategies using age reduction methods. The chosen maintenance strategy is the optimal preventive maintenance (PM) policy, which is expected to minimize total costs. In addition, the proposed strategy is expected to improve equipment performance as measured by the Overall Equipment Effectiveness (OEE). PM is done to reduce failure and avoid penalties as a result of failure. Meanwhile, Corrective Maintenance (CM) is carried out by the lessor (equipment owner) to correct the failure with minimal repairs in the lease period. In the present study, an efficient algorithm was developed to obtain an optimal PM policy and a closed form solution was obtained for case where the lifetime distribution of the equipment was Weibull. The total cost expected to use the optimal PM policy under the proposed maintenance scheme was then compared with the performance of other policies under various maintenance schemes through numerical examples.

Based on the theoretical model, a numerical method is usually necessary for obtaining the optimal preventive maintenance (PM) policy for a deteriorating system since the theoretical model becomes complicated when the system’s hazard rate function is changed after each PM. It makes the application of the theoretical model not suitable for real cases. Moreover, the theoretical model assumes using infinite time span to obtain the long-term expected number of failures. Yet, in reality, the deteriorating systems always have a finite life time. Hence, an optimal solution might not be resulted as compared to the infinite time span. Therefore, we consider using the simulation method to obtain a range of the near-optimal PM policy. The critical step of the simulation method for obtaining a near-optimal PM policy is the generation of the random variates (RV). In this research, three methods are developed to generate the required RVs of the time-between-failures (TBF) for the finite-time-span preventive maintenance model with age reduction effect. It is found that there are no significant differences among three proposed RV generating methods when comparing the dispersion of the generated RV’s. However, the rejection method is the simplest method for obtaining the near-optimal PM policies. Examples of the near-optimal PM policies are also presented in this paper.

Deep multi-agent reinforcement learning for multi-level preventive maintenance in manufacturing systems

This paper proposes a multi-phase preventive maintenance (PM) policy for leased equipment by combining the advantages of both periodic PM and sequential PM. The lease period of the equipment is divided into multiple PM phases. The PM activities within each phase are performed periodically with the convenience of implementation, while the frequency of PM for each phase is different and it gives a gradual increase because of the imperfect effect of PM. A multi-phase PM model is built up based on the age reduction method for imperfect PM with the penalty for equipment failures and overtime of repair involved. The optimal PM intervals for every PM phases are achieved by minimising the cumulative maintenance cost throughout the lease period from the perspective of the lessor. Numerical example shows that the cumulative maintenance cost under the proposed multi-phase PM policy is lower than that under periodic PM policy.

Joint optimization of [formula omitted] control chart and preventive maintenance policies: A discrete-time Markov chain approach