Optimised approach of feature selection based on genetic and binary state transition algorithm in the classification of bearing fault in BLDC motor

Abstract

This study represents an effective approach for detection and classification of bearing faults in brushless DC (BLDC) motors based on hall-sensor signal analysis. The envelope analysis and Hilbert–Huang transform are used to extract features from the time and frequency domains of each signal. A new feature selection technique is proposed based on the combination of the genetic algorithm strength and the advantage of the binary state transition algorithm. The genetic algorithm explores search space through cross-over operator while the binary state transition algorithm is based on four special transformation operators in the local exploitation capabilities. The artificial neural network and support vector machine are used as the classifier. Each model is separately analysed and compared, leading to a high possibility to distinguish the bearing faults.