No abstract

No abstract

Time domain features are employed for detection and identification of rolling element bearing faults in rotating machinery. Only five features with simple calculation are selected as features extracted directly from the original time domain vibration signals or preprocessed time domain vibration components. Three preprocessing techniques including high and band pass filtration, wavelet package transform (WPT) and envelope analysis are researched to achieve time domain features carrying the important diagnostic information of bearing conditions. An optimized artificial neural network (ANN) with rapid learning algorithm is designed and classification is performed using the ANN combined with time domain features. The model was evaluated on vibration data recorded using two accelerometers mounted on an induction motor housing subjected to a number of single point defects. The results demonstrate the proposed model is capable of high precision, fast processing and time savings in identification of bearing faults.

No abstract

No abstract