An Improved Convolutional-Neural-Network-Based Fault Diagnosis Method for the Rotor–Journal Bearings System

Abstract

More layers in a convolution neural network (CNN) means more computational burden and longer training time, resulting in poor performance of pattern recognition. In this work, a simplified global information fusion convolution neural network (SGIF-CNN) is proposed to improve computational efficiency and diagnostic accuracy. In the improved CNN architecture, the feature maps of all the convolutional and pooling layers are globally convoluted into a corresponding one-dimensional feature sequence, and then all the feature sequences are concatenated into the fully connected layer. On this basis, this paper further proposes a novel fault diagnosis method for a rotor–journal bearing system based on SGIF-CNN. Firstly, the time-frequency distributions of samples are obtained using the Adaptive Optimal-Kernel Time–Frequency Representation algorithm (AOK-TFR). Secondly, the time–frequency diagrams of the training samples are utilized to train the SGIF-CNN model using a shallow information fusion method, and the trained SGIF-CNN model can be tested using the time–frequency diagrams of the testing samples. Finally, the trained SGIF-CNN model is transplanted to the equipment’s online monitoring system to monitor the equipment’s operating conditions in real time. The proposed method is verified using the data from a rotor test rig and an ultra-scale air separator, and the analysis results show that the proposed SGIF-CNN improves the computing efficiency compared to the traditional CNN while ensuring the accuracy of the fault diagnosis.

Efficiency of diagnosing the condition of rolling bearings in real time

“…Study [6] proves that the efficiency of diagnosing a rotor journal malfunction increases with the use of the simplified global information fusion convolution neural network (SGIF-CNN). The idea of using SGIF-CNN can be good only when identifying unchangeable fault signs.…”

Bearing Fault Diagnosis Based on Image Information Fusion and Vision Transformer Transfer Learning Model

“…The image was trained in the LeNet-5 model based on CNN, and the proposed method which was tested on three famous datasets, including motor bearing dataset, self-priming centrifugal pump dataset, and axial piston hydraulic pump dataset, and had achieved a prediction accuracy of 99.79%, 99.481%, and 100%, respectively. Luo et al [17] utilized the Adaptive Optimal-Kernel Time-Frequency Representation (AOK-TFR) algorithm to transform time series into time-frequency spectrogram representations, thereby simultaneously capturing signal characteristics from both time and frequency domains.…”