基于GCNN的滚动轴承故障诊断

Fault Diagnosis of Rolling Bearings Based on GCNN

随着智能制造的快速发展,基于数据驱动的故障诊断方法逐渐成为轴承状态监测的研究重点。针对传统滚动轴承故障诊断方法所存在的特征提取和特征选择复杂且效果不佳的问题,提出一种基于卷积神经网络智能诊断算法。该算法首先利用重叠采样方法扩充数据集,再采用格拉姆角场方法将原始时域信号转化为二维图像;然后构建一个具有4个卷积层的卷积神经网络,将扩充后的数据集输入卷积神经网络进行轴承故障分类。仿真结果表明,基于格拉姆角场法和卷积神经网络的滚动轴承故障诊断方法准确率达到99.73%,高于基于传统的机器学习及同类型深度神经网络的故障诊断方法,可准确实现滚动轴承故障状态识别和分类,具有一定的应用前景。

With the rapid development of intelligent manufacturing,data-driven fault diagnosis method has gradually become the research focus of bearing condition monitoring.However,traditional rolling bearing fault diagnosis methods have the problems of complex feature extraction and feature selection and poor efficiency.This paper proposes a convolutional neural network intelligent diagnosis algorithm to solve these problems.Firstly,the overlapping sampling method is used to expand the data set,and the Gramian angle field method is used to transform the original time domain signal into a two-dimensional image.Then,a convolutional neural network with four convolution layers is constructed,and the expanded data set is input into the convolutional neural network for bearing fault diagnosis and classification.The simulation results show that the rolling bearing fault diagnosis method based on Gramian angle field method and convolution neural network can achieve the accuracy rate of 99.73%,which is higher than that of the traditional machine learning and the same type of deep neural network fault diagnosis method.It can accurately realize the identification and classification of rolling bearing fault states,and has some application prospect.