Fault Diagnosis of Rolling Bearings Under Variable Working Conditions Based on EOS-CNN

Abstract

Under time-varying working conditions, the vibration signals of rolling bearings have problems such as amplitude changes, pulsating impact intervals, non-constant sampling phase, and signal noise pollution. These factors making it difficult for data-driven methods to learn the feature information of the signals and achieve intelligent diagnosis of rolling bearings under variable working conditions. To address this issue, a rolling bearing fault diagnosis method based on envelope order spectrum and convolutional neural network (EOS-CNN) is proposed in this paper. First, an improved variational mode decomposition (VMD) method is used to decompose the original signal to obtain the optimal Intrinsic mode functions (IMFs). Then, the EOS features of IMFs are extracted using the Hilbert order transform (HOT). Finally, a one-dimensional convolutional neural network (1D CNN)-based classifier is employed to classify the extracted features and achieve intelligent diagnosis of rolling bearing faults. The proposed method is verified by publicly available variable speed bearing dataset from Ottawa University. The experiments results show that the diagnostic accuracy of the proposed method with envelope order analysis-based fault features as input is much higher than that of the original signal as input, demonstrating the effectiveness of the proposed method.