Improving the performance of univariate control charts for abnormal detection and classification

Abstract

Bearing failures in rotating machinery can cause machine breakdown and economical loss, if no effective actions are taken on time. Therefore, it is of prime importance to detect accurately the presence of faults, especially at their early stage, to prevent sequent damage and reduce costly downtime. The machinery fault diagnosis follows a roadmap of data acquisition, feature extraction and diagnostic decision making, in which mechanical vibration fault feature extraction is the foundation and the key to obtain an accurate diagnostic result. A challenge in this area is the selection of the most sensitive features for various types of fault, especially when the characteristics of failures are difficult to be extracted. Thus, a plethora of complex data-driven fault diagnosis methods are fed by prominent features, which are extracted and reduced through traditional or modern algorithms. Since most of the available datasets are captured during normal operating conditions, the last decade a number of novelty detection methods, able to work when only normal data are available, have been developed. In this study, a hybrid method combining univariate control charts and a feature extraction scheme is introduced focusing towards an abnormal change detection and classification, under the assumption that measurements under normal operating conditions of the machinery are available. The feature extraction method integrates the morphological operators and the Morlet wavelets. The effectiveness of the proposed methodology is validated on two different experimental cases with bearing faults, demonstrating that the proposed approach can improve the fault detection and classification performance of conventional control charts.

A Review of Early Fault Diagnosis Approaches and Their Applications in Rotating Machinery

“…Morsy et al [111] apply the optimal Morlet wavelet filter to preprocess the signal, and envelope detection was used to identify bearing failures at early stage. Yiakopoulos [112] proposed a feature extraction method combining the morphological analysis and complex shifted Morlet wavelets for bearing fault diagnosis. In [113], the signal was filtered using a band-pass filter determined by a Morlet wavelet which parameters were optimized based on maximum kurtosis.…”

“…Fan et al [95] Wavelet transform He et al [96] Wavelet transform Cui et al [97] Wavelet transform + time-frequency analysis + blind source Separation theory Morsy et al [111] Morlet wavelet Filter + envelope detection Yiakopoulos [112] Morphological + Complex Shifted Morlet Wavelets. Cui et al [98] High-frequency characteristics + self-adaptive wavelet de-noising Wang et al [114] Complex Morlet wavelet coefficients + sparsity measurement Tse et al [109] Wavelet transform + envelope analysis Wang et al [99] Adaptive wavelet stripping algorithm Morsy et al [113] Maximum Kurtosis + Morlet wavelet Combet et al [100] Wavelet bicoherence Moumene et al [101] Wavelets multiresolution analysis + the high-frequency resonance Fan et al [105] Discrete wavelet transform Karuppaiah et al [108] HAAR wavelet Rahman et al [106] Discrete wavelet transform Rangel-Magdaleno et al [107] Discrete wavelet transform + motor current signature analysis Chen et al [102] Adaptive redundant multiwavelet packet He et al [103] Adaptive multiwavelet Yang et al [110] EMD + autocorrelation de-noising + wavelet package decomposition Li et al [104] Intrinsic character-scale decomposition + tunable Q-factor wavelet transform.…”

Research on Fault Diagnosis of Flexible Material R2R Manufacturing System Based on Quality Control Chart and SoV

“…, where 2 indicates the alarm probability of each control chart at the beginning of the fault when = , 3 indicates probability of alarm of each control chart when > , the fault continues to occur, and the mean value of the fault source changes while the covariance matrix does not change.…”

“…The fault diagnosis method based on the quality control chart classifies the various patterns of control charts from processing quality data, establishes an abnormal pattern set and a fault set, and correlates the abnormal pattern set and the fault set in order to diagnose the fault source of the manufacturing system [2]. The existing quality control chart fault diagnosis methods are univariate control chart [3], multivariate control chart [4], regression adjustment control chart [5], and so on. However, when using these methods to monitor multistation systems, the control charts have a high false alarm rate.…”