| Popis: |
Modeling and analysis of repairable systems have drawn a lot of attention in reliability and maintenance area. Earlier studies and results in this field usually assume that a system after each repair is as same as new (perfect repair/maintenance) or as same as old (minimal repair/maintenance). These two assumptions are often found very limited uses in practical applications, most repair activities may realistically result in a complicated intermediate one (general repair or imperfect repair/maintenance). Recently, statisticians and reliability engineers start to focus more on this type of repairable systems where repair actions do not bring the system to an as same as new condition but rather bring the state of a failed system to a level that is somewhere between new and the status prior to failure, and propose various models. However, as Guo, Ascher and Love noticed, too much attention is paid to the invention of new models, with little thought; it seems, as to their applicability. Too little attention is paid to data collection and considering the usefulness of models for solving real reliability problems. To our best knowledge, these models are difficult to be used to solve engineering problems either because of the strong assumptions or the model complexity. In this paper, we propose a practical model which is based on Proportional Intensity (PI) Model and virtual time concept, and explore the tools to analyze general repairable systems. The proposed model can incorporate the time trends, proportional failure intensity as well as the cumulative repair effects. More important, unlike the current models, the closed forms for all the reliability metrics can easily be obtained by solving differential equations, thus the reliability engineers can develop the Fisher Information Matrix or Likelihood Ratio confidence bounds based on the closed form expressions. The practical use of the proposed method is demonstrated by two real case studies. The results show that our proposed method is a very promising, efficient and practical approach with the potential of becoming very useful in industry and of leading to further generalization of repairable systems analysis |
