A Novel Rolling Bearing Defect Detection Method Based on Bispectrum Analysis and Cloud Model-Improved EEMD

Abstract

Mechanical signals are not only disturbed by Gaussian noise, but also by non-Gaussian noise. These Gaussian noise and non-Gaussian noise have gravely impeded detecting of rolling bearing defects using traditional methods. In this context, the paper develops a novel detection method for rolling bearing, which combines bispectrum analysis with an improved ensemble empirical mode decomposition (EEMD). To effectively eliminate Gaussian noise in the signal, bispectrum analysis is adopted. In order to effectively reduce non-Gaussian noise, a cloud model-improved EEMD is proposed, where the cloud model is introduced to restrain the mode mixing phenomenon. Then a rolling bearing defect detection plan based on the proposed method is put forward. From theoretical analysis and experimental verification, it is demonstrated that the proposed method has superior performance in reducing multiple background noise. Furthermore, compared with other three methods, the results show that the proposed method can detect the defect of rolling bearings more effectively. INDEX TERMS Bispectrum analysis, cloud model, defect detection, ensemble empirical mode decomposition, mode mixing, rolling bearing.

Multi-Scale Sample Entropy-Based Energy Moment Features Applied to Fault Classification

“…We first fix the Embedded Dimension m = 2 to study the impact of the Tolerance r on the performance of the three feature extractors including ApprEn, SampEn and M-SSampEn-EM, measured by the averaged square distances Dt. The number of scale factor nSf in multi-scale entropy analysis and the sampling interval △t in EM analysis are fixed as nSf = [8,32] and △t = 4, respectively. The r value changes from 0.01 to 0.30.…”

“…We study spatial diversity performance of the M-SSampEn-EM feature extractor with different m values. The m value is changed from 1 to 6, fixing the other parameters as rmax = [0.04, 0.03], nSf = [8,32], and △t = 4. The change curves of Dt-m are described in figure 5 with a black-diamond line and a black-asterisk line respectively.…”

“…The change curves of Dt-m are described in figure 5 with a black-diamond line and a black-asterisk line respectively. We also fix the parameters as rmax = [0.04, 0.03] and nSf = [8,32], and change the △t value from 1 to 8. The featureextraction results by the M-SSampEn-EM feature extractor show that the Dt values are always [0.1293, 0.1197], implying the parameter △t has no impact on the performance of the M-SSampEn-EM feature extractor.…”

“…It is also a noise-assisted data analysis method, which can restrain the mode mixing phenomenon caused by the classical EMD by adding white noise into the signal. However, the quality of feature extraction will be affected by some false IMFs generated in the decomposition of complex signals by either EMD or EEMD [8].…”

A Multitask-Aided Transfer Learning-Based Diagnostic Framework for Bearings under Inconsistent Working Conditions

“…Therefore, to tackle these issues caused by significant variations in the frequency and spectrum amplitudes [ 42 , 44 ], higher-order analysis is considered by using the bispectrum of the signals. The bispectrum is mainly a third-order, spectrum-based approach that searches for non-linear interactions from given signals under inconsistent working conditions; this is done by preserving the phase information while eliminating the Gaussian noise [ 45 ]. The insights of bispectrum procedures are given below: First, a discrete signal is needed, which is defined as follows: The discrete coefficients of this signal can be expressed as: The third-order autocorrelation coefficients can be calculated as: Now, the bispectrum of can be expressed as: …”