Machine Fault Diagnosis Using IIoT, IWSNs, HHT, and SVM

Abstract

Mechanical equipment fault signals are mostly nonlinear and non-stationary signals. Processing these signals by Fourier transform and wavelet transform usually cannot obtain desired fault diagnosis results. In this paper, a machine fault diagnosis method based on Industrial Internet of Things (IIoT), industrial wireless sensor networks (IWSNs), Hilbert-Huang transform (HHT), and support vector machine (SVM) is proposed, in which HHT and SVM are used for fault feature extraction and fault diagnosis respectively. The proposed fault diagnosis approach by SVM is implemented and tested on the IWSN sensor node, while the fault feature extraction method using HHT is verified by MATLAB simulation. The effectiveness of the presented approach is evaluated by a set of experiments using bearing vibration data. The result indicates that the fault diagnosis accuracy of the presented method reaches 100% for four machine working conditions and 92% for five working conditions.