Rotating machinery has been widely used in the fields of machine tools, electric power, railway, ship, automobile, petrochemistry, etc. Owing to the severe environment and the complex working condition, the key rotating components including the gear, bearing, and rotor, are easily subject to failure, such as cracking, wear, pitting, spalling, fracture, scratch, etc. If early minor faults are not diagnosed timely, they will deteriorate rapidly and may lead to a halt of the whole rotating machinery, and even catastrophic economic losses and casualties. To keep the machine in safe and reliable operation, it is necessary to detect or predict faults as early as possible.

In recent years, fault detection and life prediction for rotating machineries have received a lot of attention. A number of methods, such as wavelet transform, empirical mode decomposition, variational mode decomposition, spectral kurtosis, time-frequency distribution, sparse representation, random forest, Markov model, support vector machine, manifold learning, particle filter, deep learning, transfer learning, etc., have been proposed to deal with these problems. Most studies have focused on fault feature extraction and fault classification using the acquired vibration signals, and there are relatively few studies on fault prognosis. Due to the variable working condition and the complex structure of such rotating machineries as aero-engines, high speed rail, and wind turbines, there are still great challenges in accurately detecting the incipient faults and predicting the long-term life of rotating machinery.

The aim of this Special Issue is to facilitate the development of fault prognosis and life prediction approaches based on vibration data for varied rotating machineries. This Special Issue hopes to attract studies including weak fault feature extraction based on various signal processing methods, fault diagnosis under variable working condition, intelligent fault diagnosis by machine learning, transfer fault recognition by domain adaption models, data-driven life prediction based on deep learning, hybrid life prediction based on both vibration data and physical model, etc. Original research and review articles are welcome.

Potential topics include but are not limited to the following:

• Sparse representation for fault feature extraction
• Fault feature extraction based on varied mode decomposition algorithms
• Fault feature extraction based on varied wavelet transforms
• Fault feature extraction under variable working condition
• Transient feature extraction by new impact indicators
• Time-frequency distributions of vibration signals for fault diagnosis
• Intelligent fault diagnosis by support vector machine
• Intelligent fault diagnosis by random forest
• Intelligent fault diagnosis by various deep neural networks
• Transfer fault diagnosis by various domain adaption models
• Vibration data-driven life prediction based deep learning
• Hybrid life prediction based on both vibration data and physical models