Remaining useful life prediction of degrading systems subjected to imperfect maintenance: Application to draught fans
Remaining useful life prediction of degrading systems subjected to imperfect maintenance: Application to draught fans
复制标题
维护不完善的退化系统的剩余使用寿命预测：在风机中的应用
DOI：
10.1016/j.ymssp.2017.08.016
复制
发表时间：
2018-02-01
影响因子：
8.4
通讯作者：
Zio, Enrico
中科院分区：
文献类型：
--
作者：
Wang, Zhao-Qiang;Hu, Chang-Hua;Zio, Enrico
目前的退化建模和剩余使用寿命预测研究共享一个共同的假设，即退化系统没有得到完美的维护或维护(即，达到新的状态)。本文研究的问题是如何对退化过程进行建模和预测退化系统在不完全维护活动下的剩余使用寿命，从而将退化系统的健康状况恢复到新旧之间的任何退化水平。为此，首先提出了一种由Wiener过程驱动的非线性模型来刻画退化系统在不完全维护下的退化轨迹，其中引入了负跳跃来量化不完全维护活动对系统退化的影响。然后，通过空间尺度变换，将所构造的具有跨越恒定阈值水平的负跳跃的退化模型转换为跨越随机阈值水平的Wiener过程模型，解析地推导出剩余使用寿命的概率密度函数。为了实现该方法，利用最大似然估计方法估计退化模型中的未知参数。最后，将所提出的退化建模和剩余使用寿命预测方法应用于某钢厂机械系统中的引风机实例。结果表明，对于不完全维修的退化系统，我们提出的方法可以得到比文献中的基准模型更准确的剩余使用寿命预测。(C)2017爱思唯尔有限公司。保留所有权利。
Current degradation modeling and remaining useful life prediction studies share a common assumption that the degrading systems are not maintained or maintained perfectly (i.e., to an as-good-as new state). This paper concerns the issues of how to model the degradation process and predict the remaining useful life of degrading systems subjected to imperfect maintenance activities, which can restore the health condition of a degrading system to any degradation level between as-good-as new and as-bad-as old. Toward this end, a nonlinear model driven by Wiener process is first proposed to characterize the degradation trajectory of the degrading system subjected to imperfect maintenance, where negative jumps are incorporated to quantify the influence of imperfect maintenance activities on the system's degradation. Then, the probability density function of the remaining useful life is derived analytically by a space-scale transformation, i.e., transforming the constructed degradation model with negative jumps crossing a constant threshold level to a Wiener process model crossing a random threshold level. To implement the proposed method, unknown parameters in the degradation model are estimated by the maximum likelihood estimation method. Finally, the proposed degradation modeling and remaining useful life prediction method are applied to a practical case of draught fans belonging to a kind of mechanical systems from steel mills. The results reveal that, for a degrading system subjected to imperfect maintenance, our proposed method can obtain more accurate remaining useful life predictions than those of the benchmark model in literature. (C) 2017 Elsevier Ltd. All rights reserved.