Failure Prediction and Wear State Evaluation of Power Shift Steering Transmission

Abstract:

Within the vehicle transmission, the friction surfaces of mechanical parts were consecutively worn-out and ultimately up to the degradation failures. For assessing the wear progress effectively, wear particles should be generally monitored by measuring the element concentration through Atomic emission (AE) spectroscopy. Herein, the spectral data sampled from life-cycle test has been processed by both the Principal Component Analysis (PCA) and further Kernel Principal Component Analysis (KPCA). Results show that KPCA acts more effectively in variable-dimensions reduction due to fewer principle components and higher cumulative contributing rate. To detect the threshold point at where the wear-stage upgraded, the Fuzzy C-means clustering algorithm was applied to process the eigenvalues of principle components. Furthermore, it is demonstrated that the principle components relate to the worn-out state of friction pairs or transmission parts. In general, the introduction of KPCA has contributed to assess the wear-stage at where the machine situates and the accurate worn-out state of various transmission parts.

• Periodical: Applied Mechanics and Materials (Volume 741)
• Pages: 183-187
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.741.183
• Online since: March 2015
• Authors: Yong Liu, Biao Ma, Yu Yan, Chang Song Zheng*
• Keywords: Failure Prediction, Oil Analysis, Power-Shift Steering Transmission, Wear State

References

[1] Toms LA and Toms AM. Machinery oil analysis-a guide for maintenance managers, supervisors and technicians [M]. Park Ride: Society of Tribologists and Lubrication Engineers, (2008).

[2] Valis D, Kouchy M and Zak L. On approaches for non-direct determination of system deterioration [J]. Eksploatacja i Niezawodnosc-Maintenance and Reliability. 2012, 14(1): 33-41.

[3] Meeker WQ, Escobar LA. Statistical methods for reliability data [M]. Chichester: John Willey and Son, (1998).

[4] Nikulin et al. Advances in degradation modeling, statistics for industry and technology [M]. Boston: Birkhauser, (2010).

[5] Makis Viliam, Wu Jianmou, Gao yan. An application of DPCA to oil data for CBM modeling [J]. European Journal of Operational Research, 2006, 174(1): 112-123.

DOI: 10.1016/j.ejor.2005.03.010

[6] Zanoli S. M, Astolfi G, Barboni L. FDI of process faults based on PCA and cluster analysis [J]. IEEE Conference on Control and Fault-Tolerant Systems. 2010, 197-202.

DOI: 10.1109/systol.2010.5676023

[7] Mia Hubert, Peter J. R, Karlien V.B. ROBPCA: a New Approach to Robust Principal Component Analysis[M]. 2003, 100-123.

[8] P J Rousseeuw, Van D.K. A Fast Algorithm for the Minimum Covariance Determinant Estimator [J]. Technometrics, 1999, 41(3): 25-31.

DOI: 10.1080/00401706.1999.10485670

[9] Gyeongyong Heo, Paul Gader, Hichem Frigui. RKF-PCA: Robust kernel fuzzy PCA [J]. Neural Networks, 2009, 22: 138~140.

DOI: 10.1016/j.neunet.2009.06.013

[10] Uwe Kruger, Zhang Jun-ping, Xie Lei. Developments and Applications of Nonlinear Principal Component Analysis-a Review[M]. 2008, 100-120.

[11] Heng Lian. Bayesian Nonlinear Principal Component Analysis Using Random Fields [J]. Pattern Analysis and Machine Intelligence. 2009, 4(31): 749~751.

DOI: 10.1109/tpami.2008.212

[12] Palomino L. V, Steffen Jr. Valder, Finzi R.M. Fuzzy cluster analysis methods applied to impedance based structural health monitoring for damage classification [J]. Conference Proceeding s of the Society for Experimental Mechanics Series. 2012, 6: 205-212.

DOI: 10.1007/978-1-4614-2419-2_18