双时间尺度下的设备随机退化建模与剩余寿命预测方法

发布时间：2018-04-13 01:39

本文选题：退化建模 + 剩余寿命预测　；参考：《自动化学报》2017年10期

【摘要】：基于退化建模的剩余寿命预测(Remaining useful life,RUL)是当前可靠性领域研究的热点.现有的退化模型都是针对单个时间尺度下的退化设备,缺少对设备性能变化与多个时间尺度相关的退化建模与剩余寿命预测方法.鉴于此,本文基于Wiener过程提出了一种双时间尺度随机退化建模与剩余寿命预测方法,用随机比例系数描述不同时间尺度之间的不确定关系,推导出了首达时间意义下设备的双时间尺度剩余寿命分布,讨论了其与基于单时间尺度退化模型得到的剩余寿命分布之间的关系,并给出了基于历史退化数据的未知参数极大似然估计方法.最后,将所提方法应用到惯性平台关键器件陀螺仪的退化建模与剩余寿命预测中,验证了方法的有效性.
[Abstract]:Residual life prediction based on degradation modeling is a hot topic in the field of reliability.The existing degradation models are all for the degradation equipment under a single time scale, and there is a lack of degradation modeling and residual life prediction methods related to the performance changes of the equipment and multiple time scales.In view of this, based on the Wiener process, this paper proposes a dual time scale stochastic degradation modeling and residual life prediction method. The stochastic proportional coefficient is used to describe the uncertain relationship between different time scales.In this paper, the dual time scale residual life distribution of the equipment in the sense of first arrival time is derived, and the relationship between the residual life distribution and the residual life distribution based on the single time scale degradation model is discussed.A method of maximum likelihood estimation of unknown parameters based on historical degradation data is presented.Finally, the proposed method is applied to the degradation modeling and residual life prediction of the key devices of inertial platform gyroscope, and the validity of the method is verified.
【作者单位】：火箭军工程大学控制工程系;火箭军工程大学信息工程系;
【基金】：国家自然科学基金(61773386,61374126,61473094,61573365,61573366) 中国科协青年人才托举工程(2016QNRC001) 陕西省自然科学基金(2015JQ6235)资助~~
【分类号】：O213.2

【相似文献】