Selective Maintenance under Uncertainty Prof. Yu LIU, Tao JIANG and Yi-Ming CHEN (University of Electronic Science and Technology of China)
Due to limited maintenance resources, selective maintenance aims at seeking the optimal maintenance actions for components during maintenance break. Traditional selection maintenance strategies are oftentimes based on two assumptions: (1) all components’ health status can be perfectly inspected; (2) maintenance time of various actions and maintenance duration are precisely known. Obviously, these two assumption are not always hold in engineering practices.
In this research, by considering the imperfection of inspections which will introduce uncertainty to components’ states and effective ages, a new robust selective maintenance strategy is proposed. Firstly, the posterior probability distributions of components’ states and effective ages can be inferred by the Bayesian method. Secondly, the uncertainty of system reliability can be characterized by the expectation and variance of components’ reliability measures. Thirdly, a multi-objective robust selective maintenance model is proposed with the aim of maximizing the expectation and minimizing the variance of system reliability. As demonstrated by numerical studies and examples, the proposed approach can improve system robustness effectively.
On the other hand, by taking account of the uncertainty of maintenance time and duration, a selective maintenance strategy with considering the sequences of maintenance tasks is developed. The possible numbers of maintenance actions to be completed during an uncertain maintenance break can be formulated as a multi-dimensional convolution problem, and it is effectively computed by the saddlepoint approximation method. The ant colony algorithm is used to solve the resulting optimization problem. The proposed method is demonstrated by a set of illustrative examples.
On Determination of the Non-periodic Preventive Maintenance Scheduling with the Failure Rate Threshold for Repairable System Mr. Juhyun LEE, Jihyun PARK and Suneung AHN (Hanyang University)
Determination of a preventive maintenance scheduling is regarded as a key part in manufacturing system to maintain the equipment in good condition. In practice, many preventive maintenance policies is used in manufacturing system to reduce the unexpected failures and increase sustainability of system. In this paper, the failure rate of system is used as a condition variable and a decision variable, and preventive maintenance policy is then developed for minimizing the expected maintenance cost rate. The imperfect preventive maintenance activities of the proposed models are modeled via the arithmetic reduction model which uses the age reduction factor or the hazard reduction factor. The results of the numerical example shows that the model based on age reduction can not only extend the system lifetime but also reduce the expected maintenance cost rate although the preventive maintenance cost is significantly higher than those of the model based on hazard reduction.
Functional Reliability Prediction of Pyrotechnic Separation Device Using Response Surface Model Mr. Dong-seong KIM, Seung-Gyo JANG and Byungtae RYU (Agency for Defence Development)
Pyrotechnic mechanical device (PMD) is used in military and aerospace application like separation device because of this benefit such as fast operating time, simple structure, and light weight. However, PMD is one-shot device and it cannot be reused and repaired. Thus, reliability prediction of the PMD is the essential ingredient of design.
This paper introduces an approach to predict functional reliability of pyrotechnic separation device. The overall procedure consists of the following steps: 1) design parameters affecting the pyrotechnic separation device are identified. 2) design of experiment is defined by sensitivity analysis about parameters. 3) response surface model is constructed using polynomial equation. Later, response surface model will be commensurate with testing results.