Bearing fault classification using ANN-based Hilbert footprint analysis

Ball bearings are considered as a critical element in various mechanical systems. Vibration signal analysis is very effective method for finding bearing fault. Accelerometers are used to capture the multi-component vibration signal generated in the machine when it is in use. Various methods based on empirical mode decomposition (EMD) have been used for ball bearing fault diagnosis. EMD method usually suffered from the boundary distortion of intrinsic mode function. Classification of ball bearing fault is one of the challenging tasks in the field of mechanical systems. Various classification schemes such as support vector machine (SVM), K-means clustering, extreme learning machine (ELM) have been used for the classification of ball bearing fault. In this study, footprint analysis of Hilbert transform along with the neural network has been done for ball bearing fault analysis. A comparative analysis of the proposed research study has been done with available methods such as SVM and ELM. A high fault classification accuracy has been achieved using the proposed method for detection of ball bearing fault.