作者：任子强, 司小胜, 胡昌华, 王玺, 裴洪

作者单位：火箭军工程大学导弹工程学院, 陕西 西安 710025

关键词：复合健康指标;非线性退化模型;贝叶斯参数更新;剩余寿命预测

摘要：

针对目前基于单个传感器剩余寿命预测方法存在预测精度不高的问题，该文提出一种融合多源传感器数据的非线性退化建模与剩余寿命预测方法。该方法包括复合健康指标的构建、模型参数的估计和传感器融合系数的确定，在确定融合系数后，结合设备历史寿命数据与实时监测数据，利用Bayesian参数更新公式推导出设备的剩余寿命概率分布，实现设备的剩余寿命在线预测。最后通过由商用模块化航空推进系统仿真生成的发动机退化数据集进行仿真实验，结果表明该文所提方法能够有效提高设备剩余寿命预测的准确性。

Multi-source data fusion for nonlinear degradation modeling and remaining useful life prediction
REN Ziqiang, SI Xiaosheng, HU Changhua, WANG Xi, PEI Hong
School of Missile Engineering, Rocket Force University of Engineering, Xi'an 710025, China
Abstract: Aiming at the problem that the remaining useful life prediction accuracy is not high based on the single sensor’s signals, the article proposes a remaining useful life prediction method for equipment which is modeled by nonlinear degradation model fusing multi-sensors data. The method includes the construction of a composite health indicator, estimation of model parameters, and determination of sensors’ fusion coefficients. After determining the fusion coefficient, combined with historical life data and real-time monitoring data of equipment, the Bayesian parameter update is used to derive the remaining useful life probability distribution, and then realize the online remaining useful life prediction of a equipment. Finally, the simulation experiment is carried out by the engine degradation data set generated by commercial modular aero-propulsion system simulation. The results show that the proposed method can effectively improve the accuracy of engine’s remaining useful life prediction.
Keywords: composite health indicator;nonlinear degradation model;Bayesian parameter update;remaining useful life prediction
2020, 46(2):1-8  收稿日期: 2019-05-14;收到修改稿日期: 2019-07-02
基金项目: 国家自然科学基金(61833016，61573365)
作者简介: 任子强(1995-),男,重庆市人,硕士研究生,专业方向为故障诊断与寿命预测
参考文献
[1] 喻勇, 司小胜, 胡昌华, 等. 数据驱动的可靠性评估与寿命预测研究进展:基于协变量的方法[J]. 自动化学报, 2018, 44(2):216-227
[2] 胡昌华, 马芳, 张峰, 等. 精密机电设备寿命评估方法[J]. 导航与控制, 2018, 17(1):21-33
[3] CHEHADE A, SONG C, LIU K, et al. A data-level fusion approach for degradation modeling and prognostic analysis under multiple failure modes[J]. Journal of Quality Technology, 2018, 50(2):150-165
[4] 施权, 胡昌华, 司小胜, 等. 考虑执行器性能退化的控制系统剩余寿命预测方法[J]. 自动化学报, 2019, 45(5):941-952
[5] LIU K, GEBRAEEL N, SHI J. A data-level fusion model for developing composite health indices for degradation modeling and prognostic analysis[J]. IEEE Transactions on Automation Science and Engineering, 2013, 10(3):652-664
[6] 李楷. 机电一体化系统中传感器技术的运用研究[J]. 山东工业技术, 2019(12):145
[7] LIU K, CHEHADE A, SONG C. Optimize the signal quality of the composite health index via data fusion for degradation modeling and prognostic analysis[J]. IEEE Transactions on Automation Science and Engineering, 2017, 14(3):1504-1514
[8] GEBRAEEL N. Sensory-updated residual life distributions for components with exponential degradation patterns[J]. IEEE Transactions on Automation Science and Engineering, 2006, 3(4):382-393
[9] 赵广社, 吴思思, 荣海军. 多源统计数据驱动的航空发动机剩余寿命预测方法[J]. 西安交通大学学报, 2017, 51(11):150-155
[10] MEHR C B, MCAFADDEN J A. Certain properties of gaussian processes and their first-passage times[J]. Journal of the Royal Statistical Society, 1965, 27(3):505-522
[11] COX D R, MILLER H D. The theory of stochastic processes[M]. New York, USA:Wiley, 1965:81-83.
[12] SAXENA A, GOEBEL K. C-MAPSS data set[Z]. NASA Ames Prognostics Data Repository, 2008.
[13] SI X S, WANG W, CHEN M, et al. A degradation path-dependent approach for remaining useful life estimation with an exact and closed-form solution[J]. European Journal of Operational Research, 2013, 226(1):53-66
[14] 刘竞妍, 张可, 王桂华. 综合评价中数据标准化方法比较研究[J]. 数字技术与应用, 2018, 36(6):84-85
[15] SI X S, WANG W, Hu C H, et al. Remaining useful life estimation based on a nonlinear diffusion degradation process[J]. IEEE Transactions on Reliability, 2012, 61(1):50-67
[16] 裴洪, 胡昌华, 司小胜, 等. 融合寿命数据与退化数据的剩余寿命估计方法[J]. 电光与控制, 2016, 23(9):90-95