Bearing Fault Detection Based on Maximum Likelihood Estimation and Optimized ANN Using the Bees Algorithm

Rotating machinery is a common class of machinery in industry. The root cause of faults in rotating machinery is often faulty rolling element bearings. These rolling element bearings wear out easily due to the metal-metal contacts and create faults in the outer race, inner race or balls. This paper presents an algorithm using optimized artificial neural network by the Bees Algorithm for automated diagnosis of localized faults in rolling element bearings. Maximum likelihood estimation values extracted from time-domain vibration signals, real Cepstrum, minimum phase reconstruction, discrete cosine transform, discrete Fourier transform, envelope analysis signal and the Hilbert transform are used as input features for the neural network. The proposed procedure requires only a few input features, resulting in simple preprocessing and faster training. Effectiveness of the proposed method is illustrated using the experimentally obtained bearing vibration data.

Keywords:

Fault diagnosis , MLE distributions , RBF neural network , Bees Algorithm