A New Fault Feature Extraction Method for Rotating Machinery Based on Multiple Sensors

Abstract

During the operation of rotating machinery, the vibration signals measured by sensors are the aliasing signals of various vibration sources, and they contain strong noises. Conventional signal processing methods have difficulty separating the aliasing signals, which causes great difficulties in the condition monitoring and fault diagnosis of the equipment. The principle and method of blind source separation are introduced, and it is pointed out that the blind source separation algorithm is invalid in strong pulse noise environments. In these environments, the vibration signals are first de-noised with the median filter (MF) method and the de-noised signals are separated with an improved joint approximate diagonalization of eigenmatrices (JADE) algorithm. The simulation results found here verify the effectiveness of the proposed method. Finally, the vibration signal of the hybrid rotor is effectively separated by the proposed method. A new separation approach is thus provided for vibration signals in strong pulse noise environments. extract the characteristic signal. Cyclostationary signal analysis [21,22] relies on the cyclostationary characteristics of the bearing fault impulse signal to design a filter to eliminate random noise. A Wiener filter [23,24] is used to eliminate the stationary random noise in the fault impulse signal. A wavelet transform [25,26] first decomposes the vibration signal into different frequency bands, then defines the threshold to eliminate the noise components and reconstructs the characteristic signal. A Kalman filter [27,28] first establishes the vibration state model and signal observation model of rolling bearing, then iteratively eliminates the noise in the signal. Stochastic resonance [29,30] uses noise to enhance the fault characteristic component of the bearing vibration signal. All of the above mentioned methods are universal, can effectively eliminate background noise and interference components, and extract fault signals in specific environment, but they are not suitable for complex interference situations, and especially when the interference components in vibration signals are similar to fault signals, it is difficult to distinguish them by the above method, let alone eliminating interference components and extract fault signals. Blind source separation (BSS) technology can realize the separation of multiple aliased signals [1,31]. Meanwhile, blind source separation is not affected by the time and frequency overlap of source signals, and the separated output signal will not lose the weak feature information in the source signal.So far, many effective and distinctive blind source separation algorithms have been constructed. Typical algorithms include fast fixed-point [32] algorithms, natural gradient [28] algorithms, second-order blind identification (SOBI) [33] algorithms, equivalation adaptive separation via independence (EASI) [34] algorithms, and joint approximate diagonalization of eigenmatrices (JADE) [35] algorithms. When separating the noiseless mixed signals, these ...

A New Method of Denoising of Vibration Signal and Its Application

“…According to the size of M(x) on different decomposition scales, the optimal wavelet packet basis can be determined. reshold denoising algorithm was proposed by Donoho [15,16]. e algorithm determines the threshold value and threshold processing function according to the distribution of wavelet coefficients of the signal and noise on each decomposition scale.…”

Rolling Bearing Fault Diagnosis Based on Nonlinear Underdetermined Blind Source Separation

“…Li et al [11] combined independent component analysis (ICA) with fuzzy k−nearest neighbor to diagnose the multi−faults of gears. Miao et al [12] utilized median filter and the improved joint approximate diagonalization of eigenmatrices algorithm to identify the faults of rotating machinery. However, most studies mainly focus on the BSS problems in which the number of observed signals exceeds the number of source signals, namely the overdetermined BSS (OBSS) problem.…”

Multisource Fault Signal Separation of Rotating Machinery Based on Wavelet Packet and Fast Independent Component Analysis

“…BSS plays an increasingly important role in the field of digital signal processing and has been widely used in communication [27], speech processing [28], fault diagnosis [29,30], seismic exploration [31], biomedicine [32,33], image processing [34], radar [35], and economic data analysis [36]. In blind signal separation, the typical algorithms commonly used include the fast fixed-point algorithm [37], natural gradient algorithm [38], Equivariant Adaptive Separation via Independence (EASI) algorithm [39,40], and Joint Approximation Diagonalization of Eigen-matrices (JADE) algorithm [41,42], etc. Grotas et al [43] developed the constrained maximum likelihood (ML) estimator of the Laplacian matrix for this graph BSS problem with Gaussian-distributed states.…”