基于竞争失效的综合传动剩余寿命预测
|
摘要
剩余寿命预测是机械传动系统故障诊断和健康管理的关键。车辆综合传动装置作为复杂时变的机电系统,在高速重载的运行过程中,常存在多种失效形式和失效过程,不同失效模式间竞争失效。然而,现有综合传动装置剩余寿命预测研究中,没有将劣化失效与突发失效同时考虑。鉴于此,本文中研究基于劣化失效和突发失效竞争失效模式的综合传动装置寿命预测方法。首先,针对综合传动装置的多维监测数据,采用鲁棒核主元分析与状态空间模型融合得到整机劣化程度指标。然后,采用随机过程Wiener模型建立装置的劣化失效模型,采用Weibull分布模型建立装置突发失效模型,进一步在考虑劣化失效与突发失效相关的基础上建立了综合传动装置竞争失效模型。最后,采用最大似然估计方法估计了模型的参数,实现了综合传动装置的剩余寿命预测。实验研究验证了本文方法的有效性,能够客观描述装置的可靠度变化规律,提高剩余寿命预测的准确性。
Remaining useful life(RUL) prediction is the key issue to fault diagnosis and health management of mechanical transmission system. As a complex time-varying electromechanical system, the power-shift steering transmission(PSST) often has many failure modes and processes in the running process under high speed and heavy load, and the interaction between different failure modes is competition failure. However, in current RUL prediction studies of PSST, there is no study considering the competing failure. In this paper, the RUL prediction method of PSST is presented based on competing risks of degradation failure and random failure. First, the multi-dimensional degradation monitoring data is integrated by the state space model and robust kernel principle component analysis(PCA) to obtain the degradation index of the whole machine. Then, the degradation model is established based on Wiener process, the random failure model is established based on Weibull distribution, and the competition failure model of PSST is further established considering the correlation between the degradation failure and random failure. Finally, the model's parameters are estimated by the maximum likelihood estimation(EML) method and the RUL prediction of PSST is realized. The experiments validate the proposed method, and the results indicate that the proposed method can objectively describe the reliability variation law of the device and can effectively improve the accuracy of RUL prediction.
Remaining useful life(RUL) prediction is the key issue to fault diagnosis and health management of mechanical transmission system. As a complex time-varying electromechanical system, the power-shift steering transmission(PSST) often has many failure modes and processes in the running process under high speed and heavy load, and the interaction between different failure modes is competition failure. However, in current RUL prediction studies of PSST, there is no study considering the competing failure. In this paper, the RUL prediction method of PSST is presented based on competing risks of degradation failure and random failure. First, the multi-dimensional degradation monitoring data is integrated by the state space model and robust kernel principle component analysis(PCA) to obtain the degradation index of the whole machine. Then, the degradation model is established based on Wiener process, the random failure model is established based on Weibull distribution, and the competition failure model of PSST is further established considering the correlation between the degradation failure and random failure. Finally, the model's parameters are estimated by the maximum likelihood estimation(EML) method and the RUL prediction of PSST is realized. The experiments validate the proposed method, and the results indicate that the proposed method can objectively describe the reliability variation law of the device and can effectively improve the accuracy of RUL prediction.
引文
[1] 闫书法,马彪,郑长松.基于多元劣化失效的综合传动剩余寿命预测[J].华中科技大学学报:自然科学版,2018,46(8):28-33.
[2] LEE J,WU F,ZHAO W,et al.Prognostics and health management design for rotary machinery systems—reviews,methodology and applications[J].Mechanical Systems & Signal Processing,2014,42(1):314-334.
[3] YAN S F,MA B,ZHENG C S.Degradation index construction methodology for mechanical transmission based on fusion of multispectral oil data[J].Industrial Lubrication and Tribology,2018.https://doi.org/10.1108/ILT-04-2018-0154.
[4] YAN S F,MA B,ZHENG C S.Remaining useful life prediction for power-shift steering transmission based on fusion of multiple oil spectra[J].Advances in Mechanical Engineering,2018,10(6).https://doi.org/10.1177/1687814018784201.
[5] 闫书法,马彪,郑长松,等.基于劣化数据的综合传动装置剩余寿命预测[J].北京理工大学学报,2018,38(11):1126-1133.
[6] YAN S F,MA B,ZHENG C S.Health index extracting methodology for degradation modelling and prognosis of mechanical transmissions[J].Eksploatacja i Niezawodnosc-Maintenance and Reliability,2019,21(1):137-144.
[7] 闫书法,马彪,郑长松,等.基于不确定油液光谱数据的综合传动装置剩余寿命预测[J].光谱学与光谱分析,2019,39(2):553-558.
[8] HUNG H H,YEH Y R.An iterative algorithm for robust kernel principal component analysis[J].Neurocomputing,2011,74(18):3921-3930.
[9] ICHIHASHI H,MIYAGISHI K,HONDA K.Fuzzy c-means clustering with regularization by KL information[C].Proceedings of 10th IEEE International Conference on Fuzzy Systems(ICFS).Melbourne,Australia:IEEE,2001,2:924-927.
[10] HAMM J,LEE D D.Grassmann discriminant analysis:a unifying view on subspace-based learning[C].Proceedings of the 25th International Conference on Machine Learning(ICML).Helsinki,Finland:ACM,2008:376-383.
[11] 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.
[12] CHATZIDIAMANTIS N D,SANDALIDIS H G,KARAGIANNIDIS G K,et al.Inverse gaussian modeling of turbulence-induced fading in free-space optical systems[J].Journal of Lightwave Technology,2011,29(10):1590-1596.
[13] SU C,ZHANG Y.System reliability assessment based on Wiener process and competing failure analysis[J].Journal of Southeast University (English Edition),2010,26(4):554-557.
[2] LEE J,WU F,ZHAO W,et al.Prognostics and health management design for rotary machinery systems—reviews,methodology and applications[J].Mechanical Systems & Signal Processing,2014,42(1):314-334.
[3] YAN S F,MA B,ZHENG C S.Degradation index construction methodology for mechanical transmission based on fusion of multispectral oil data[J].Industrial Lubrication and Tribology,2018.https://doi.org/10.1108/ILT-04-2018-0154.
[4] YAN S F,MA B,ZHENG C S.Remaining useful life prediction for power-shift steering transmission based on fusion of multiple oil spectra[J].Advances in Mechanical Engineering,2018,10(6).https://doi.org/10.1177/1687814018784201.
[5] 闫书法,马彪,郑长松,等.基于劣化数据的综合传动装置剩余寿命预测[J].北京理工大学学报,2018,38(11):1126-1133.
[6] YAN S F,MA B,ZHENG C S.Health index extracting methodology for degradation modelling and prognosis of mechanical transmissions[J].Eksploatacja i Niezawodnosc-Maintenance and Reliability,2019,21(1):137-144.
[7] 闫书法,马彪,郑长松,等.基于不确定油液光谱数据的综合传动装置剩余寿命预测[J].光谱学与光谱分析,2019,39(2):553-558.
[8] HUNG H H,YEH Y R.An iterative algorithm for robust kernel principal component analysis[J].Neurocomputing,2011,74(18):3921-3930.
[9] ICHIHASHI H,MIYAGISHI K,HONDA K.Fuzzy c-means clustering with regularization by KL information[C].Proceedings of 10th IEEE International Conference on Fuzzy Systems(ICFS).Melbourne,Australia:IEEE,2001,2:924-927.
[10] HAMM J,LEE D D.Grassmann discriminant analysis:a unifying view on subspace-based learning[C].Proceedings of the 25th International Conference on Machine Learning(ICML).Helsinki,Finland:ACM,2008:376-383.
[11] 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.
[12] CHATZIDIAMANTIS N D,SANDALIDIS H G,KARAGIANNIDIS G K,et al.Inverse gaussian modeling of turbulence-induced fading in free-space optical systems[J].Journal of Lightwave Technology,2011,29(10):1590-1596.
[13] SU C,ZHANG Y.System reliability assessment based on Wiener process and competing failure analysis[J].Journal of Southeast University (English Edition),2010,26(4):554-557.