Modulated Gabor filter based deep convolutional network for electrical motor bearing fault classification and diagnosis

The high applicability of the electrical motor has led to gain attention in condition monitoring diagnosis most common type fault this machine, bearing element. emergence deep neural networks (DNN) provides opportunity design a network for early with speed and without any additional feature extraction technique. However, robustness against noise some deficiencies fully capturing features are still challenging issues. To resolve problem, paper proposes one module Gabor filter based convolutional (CNN), namely (GCNN), detection classification. GCNN is modulated enhance ability temporal as well understanding spatial fewer parameters higher noises can be considered computational efficient structure. simulation results detection/classification studied on two different experimental prototypes, including case Western Reserve University (CWRU) Paderborn (Paderborn) datasets. superiority method shown by comparison accelerated CNN (ACNN), adaptive CNN, standard support vector machine (SVM), learning quantisation (LVQ), feedforward (FFNN).