An Improved Fault Diagnosis Using 1D-Convolutional Neural Network Model

• DOI: 10.3390/electronics10010059
• Publication Date: 2020-12-31T15:10:37Z
• Authors: Chih-Cheng Chen; Zhen Liu; Guangsong Yang; Chia-Chun Wu; Qiubo Ye
• Keywords: Convolution (computer science); Kernel (algebra); Dropout (neural networks)

The diagnosis of a rolling bearing for monitoring its status is critical in maintaining industrial equipment while using bearings. traditional method diagnosing faults the has low identification accuracy, which needs artificial feature extraction order to enhance accuracy. one-dimensional convolution neural network (1D-CNN) can not only diagnose accurately, but also overcome shortcomings methods. Different from machine learning and other deep models, 1D-CNN does need pre-processing data bearing’s vibration. In this paper, architecture proposed effectively improve accuracy bearing, number kernels decreases with reduction kernel size. obtains high improves generalizing ability by introducing dropout operation. experimental results show 99.2% average under single load 98.83% different loads.