Smart machine fault diagnostics based on fault specified discrete wavelet transform

Machine learning for fault analysis in rotating machinery: A comprehensive review

“…On the other hand, these rotor faults are mostly examined regarding single-fault occurrences where it is assumed that a single fault exists in the system. Although there are studies that investigated compound fault conditions [ 11 , 40 , 114 ], they considered a rotor fault that occurred at the same time as a fault of another component such as a bearing or gear.…”

“…Visual Data comprises image data obtained through high-resolution cameras, optical devices, or images derived from a signal. Visual data may include thermal images [ 32 , 114 ], x-ray images [ 143 , 144 ], acoustic signals [ 11 , 23 , 24 ], or vibration data [ 75 , 85 , 106 ] or their processed form including spectrogram [ 145 , 146 ], waveforms [ 147 ], etc that usually reflects the characteristics of the healthy and faulty behavior of a component of a machine. Acoustic Data includes the sound or noise that is produced by a rotating machine that differs when operating in a healthy and faulty condition.…”

“…Various ML techniques are practiced for different kinds of problems. The most commonly used algorithms are Deep Neural Networks (DNN) algorithms [ [1] , [2] , [3] , [4] , [5] ], Artificial Neural Networks (ANN) [ [6] , [7] , [8] ], AdaBoost (AB) [ 8 , 9 ], Random Forest (RF) [ 10 , 11 ], Support Vector Machines (SVM) [ [12] , [13] , [14] ], K Nearest Neighbors (KNN) [ 12 , 13 , 15 ], and Decision Trees (DT) [ 16 , 17 ]. As the concept of the Fourth Industrial Revolution or Industry 4.0 is introduced, the use of ML becomes even more critical, especially for the industry itself.…”

“…To develop Intelligent Fault Diagnosis and Prognosis (IFDP) model, researchers need to consider suitable data sources, preprocessing techniques, machine learning algorithms, and optimization algorithms. For this purpose, they deemed vibration [ 11 , 14 , 15 , 21 , 22 ], acoustic [ 11 , 23 , 24 ], thermal [ 13 ], current [ 6 , 7 , 9 , 25 , 26 ], pressure [ 27 ], and other characteristic data [ [27] , [28] , [29] ] as the main source for IFDP of rotating machines. Afterwards, the distinctive features are extracted by employing feature extraction methods such as statistical feature extraction [ [30] , [31] , [32] ], Fourier Transform [ [33] , [34] , [35] ], Wavelet Transform [ [36] , [37] , [38] ], Empirical Mode Decomposition [ 28 , 39 , 40 ] or other techniques [ 6 , 7 , 41 , 42 ].…”

Generation of high-dimensional vibration signal and its application in fault diagnosis

“…However, designing appropriate health indicators and setting suitable thresholds are challenging under complex conditions, thereby hindering the generalization performance of the method [6]. Intelligent fault diagnosis models have emerged as promising alternatives [7]. These models can adaptively extract healthstatus sensitive features from intricate monitoring data and establish correlations with the health status of the equipment [8].…”