报告摘要：
This seminar will present some work carried out at the University of Ottawa in the area of machine fault detection. The focus will be on two methods: a) a joint time-invariant wavelet transform and kurtosis method for on-line oil-debris detection, and b) a differentiation based method for bearing fault detection.

On-line oil debris sensors are important devices for detection of machinery failures. However, two key issues are yet to be addressed to make use of the existing oil debris sensors more effectively: the responsiveness to early machine failures and false alarms. The responsiveness level depends on the size of the debris that can be detected by an oil debris sensor. The detectable particle size in turn is mainly limited by the background noise. The false alarms are often caused by spurious impulses such as vibration-like signals. The challenge of improving the responsiveness and reducing false alarms lies in the very weak particle signals and their similarity to spurious signals. The joint time-invariant wavelet transform and kurtosis analysis method is proposed to address the two issues simultaneously. Our test results have demonstrated that the proposed method can effectively detect very weak particle signals buried in strong background noise and eliminate vibration-like spurious signals.

Bearing faults can lead to malfunction and ultimately complete stall of many machines. The conventional high frequency resonance (HFR) method has been commonly used for bearing fault detection. However, it is often very difficult to obtain and calibrate bandpass filter parameters, i.e., the center frequency and bandwidth, the key to the success of the HFR method. This inevitably undermines the usefulness of the conventional HFR technique. To avoid such difficulties, we propose a differentiation method to enhance the fault detectability, thereby avoiding the difficulty present in the HFR method. It is shown that the iterative application of a differentiation step can boost the relative strength of the impulsive faulty bearing signal component with respect to the vibration interferences. This preserves the effectiveness of amplitude demodulation and hence leads to more accurate fault detection. The proposed approaches are evaluated on simulated signals and experimental data acquired from faulty bearings.

报告人简介：
Ming Liang is a professor of Mechanical Engineering, University of Ottawa, Canada since 2000. Dr. Liang worked as an associate professor (1996-2000) and assistant professor (1991-1996) in the same department. He received his PhD in Industrial Engineering from University of Windsor, Canada in 1991 and his M. Sc. and B.Sc. from Northeastern University, China in 1984 and 1982 respectively. His research interests include machine and process fault detection and diagnosis, machining control and monitoring, and manufacturing planning. He has collaborated with the various organizations including National Research Council (NRC) of Canada, Okuma and GasTOPS, Ltd. Dr. Liang has been serving as an Editorial Board Member of the IIE Transactions since 1996 and is an active reviewer for more than 20 professional journals (including IEEE Transactions, ASME Transactions, MSSP, JSV, MST, and IIE Transactions) and various granting agencies. Dr. Liang has served in the Grant Selection Committee (GSC) of Natural Sciences and Engineering Research Council of Canada (NSERC) (2007-2010), and NSERC Scholarships and Fellowships Selection Committee (2003-2005). He is a recipient of the Joseph Whitworth Award of the IMechE (2005) and has been listed in Who’s Who in the World (2007, 2009, 2010). Dr. Liang is a Registered Professional Engineer in the Province of Ontario, Canada and a senior member of IIE.