Blind source separation for vibration-based diagnostics of rotorcraft bearings

Abstract

The vibration signals from sensors monitoring the activity of individual bearings in a power train unit may be linear instantaneous mixtures of vibrations generated by various dynamic components. Generally, an exact physical model describing the mixing process and the contribution of each dynamic component to the received sensor signal is not available. Vibration source signals from defective bearings often overlap in time and frequency, and, as such, the direct use of time- and frequency-domain methods may result in erroneous diagnostic information. This paper implements blind source separation (BSS) to demix sensor signals into correctly identifiable vibration source signals without the need of the vibration path property and sensor layout. Experimental vibration data from spalled, corroded, and healthy rotorcraft bearings are used with five representative BSS algorithms. The separation accuracy of these algorithms is then compared using various performance metrics. Results show that despite the inherent statistical dependence and near Gaussianity, it is possible to isolate vibration sources from mixed sensor signals using second- and higher-order statistics of the signals. The paper also identifies the limitations of the BSS technique and provides a remedy and recommendation for its implementation in rotorcraft bearing fault detection.

Compound faults detection of the rolling element bearing based on the optimal complex Morlet wavelet filter

“…depends on the likelihood function p y k x k j ð Þ, which is defined by the measurement function (9) and the known process n k . With this, the recurrence relations (10) and (11) form a rigorous solution framework for the Bayesian filtering approach. Nevertheless, in many cases, the dynamics systems are not linear, so it is difficult to analytically evaluate these distributions due to some high-dimensional integrals.…”

“…The operational status of the rolling element bearing often directly affects the performance of the whole machine. In recent decades, researchers continue to pay great attention to the fault diagnosis of rolling element bearings on account of their irreplaceable importance, and have proposed many different techniques 1–12 in the literature, such as high-frequency resonance technique (HFRT), 3 spectral kurtosis (SK), 6 optimal wavelet filter based method, 7 protrugram, 9 etc. Majority of achievements are carried out for the single localized defect occurred either on the surface of outer race, inner race, or the rolling element.…”

Tool wear condition monitoring based on continuous wavelet transform and blind source separation

“…These techniques can be classified into several major approaches: non-gaussianity, maximum likelihood, minimum mutual information, neural network modeling, and algebraic [34,35]. Moreover, there are several famous algorithms which are based on the algebraic approach [24] such as FastICA, AMUSE, SOBI, JADE, and COMBI. Five algorithms implemented in this study were selected from the most used in the fault diagnosis [36] presented in this paper.…”

“…BSS has become an appealing field of research with many technological applications areas such as medical, image processing, and communications. Lately, it was applied to condition monitoring of rotating machinery [21][22][23][24]. However, little has been investigated with the application of the BSS for tool wear condition monitoring.…”

Bearing damage classification using instantaneous energy density

“…The acquired raw vibration signal can be processed in time, frequency, and time-frequency domain to extract useful information related to the severity and type of bearing damage. Many signal processing techniques exist in all these domains for bearing diagnosis (McFadden and Smith, 1984a; McFadden and Smith, 1984b; Howard, 1994; Prabhakar et al., 2002; Kar and Mohanty, 2004; Yu et al., 2006; Randall, 2011; Yaqub and Loparo, 2014; Moustafa et al., 2014; Haile and Dykas, 2015).…”