航空学报 > 2020, Vol. 41 Issue (2): 223291-223291   doi: 10.7527/S1000-6893.2019.23291

基于非线性Wiener过程的航空发动机性能衰减建模与剩余寿命预测

王玺1, 胡昌华1, 任子强1, 熊薇2   

  1. 1. 火箭军工程大学 导弹工程学院, 西安 710025;
    2. 国防大学 联合勤务学院, 北京 100089
  • 收稿日期:2019-07-16 修回日期:2019-08-14 出版日期:2020-02-15 发布日期:2019-09-30
  • 通讯作者: 胡昌华 E-mail:hch_reu@sina.com
  • 基金资助:
    国家自然科学基金(61833016,61573365)

Performance degradation modeling and remaining useful life prediction for aero-engine based on nonlinear Wiener process

WANG Xi1, HU Changhua1, REN Ziqiang1, XIONG Wei2   

  1. 1. College of Missile Engineering, Rocket Force University of Engineering, Xi'an 710025, China;
    2. College of Joint Service, National Defence University, Beijing 100089, China
  • Received:2019-07-16 Revised:2019-08-14 Online:2020-02-15 Published:2019-09-30
  • Supported by:
    National Natural Science Foundation of China (61833016, 61573365)

摘要: 针对航空发动机在性能衰减过程中普遍存在的非线性和三源不确定性问题,提出了一种基于非线性Wiener过程的航空发动机性能衰减建模与剩余寿命(RUL)预测方法。首先,为解决目前大多数剩余寿命预测方法中潜在假设的局限性,即当前时刻估计的漂移系数与上一时刻漂移系数的后验估计完全相等,在状态空间模型的框架下建立了一类新的同时考虑非线性和三源不确定性的性能衰减模型,并在首达时间下推导出剩余寿命的分布。然后,针对新研发航空发动机缺乏历史数据和先验信息的问题,提出了一种基于Kalman滤波和条件期望最大化(ECM)算法的参数估计方法,使得估计的模型参数不依赖于历史数据量。同时能够在获得一个新的性能衰减数据后,实现对模型参数的自适应估计和在线更新,进而实时地更新航空发动机的剩余寿命分布。实验结果表明,本文方法可以有效地提高剩余寿命预测的准确性,能为航空发动机的维修决策提供可靠的依据。

关键词: Wiener过程, 寿命预测, 性能衰减建模, 条件期望最大化, 航空发动机

Abstract: For the nonlinearity and three-source variability of aeroengines in the performance degradation process, a performance degradation modeling and Remaining Useful Life (RUL) prediction method for aero-engines based on nonlinear Wiener process is proposed. First, in order to solve the limitations of potential hypothesis in most current RUL prediction methods, that is, the drift coefficient of the current time estimate is exactly equal to the posterior estimate of the drift coefficient of the previous time, a new class of performance degradation model considering both nonlinearity and three-source variability is established under the framework of state space model. Further, the associated RUL distribution is derived under the first hitting time. Then, for the newly developed aero-engines lacking historical data and prior information, a parameter estimation method based on the Kalman filtering and Expectation Conditional Maximization (ECM) algorithm is proposed, so that the estimated model parameters are independent of the historical data volume. After obtaining a new performance degradation data, the model parameters can be estimated adaptively so as to update the RUL distribution of aeroengines in real time. The experimental results show that the proposed method can effectively improve the accuracy of RUL prediction and provide a reliable basis for maintenance decision of aeroengines.

Key words: Wiener process, life prediction, performance degradation modeling, expectation conditional maximization, aero-engine

中图分类号: