摘要
针对一维信号作为卷积神经网络输入时无法充分利用数据间的相关信息的问题,提出GADF-CNN的轴承故障诊断模型。利用格拉姆角差域(GADF)对采集到的振动信号进行编码,可以很容易地进行角度透视,从而识别出不同时间间隔内的时间相关性并生产相应特征图,之后将其输入卷积神经网络(CNN)自适应的完成滚动轴承故障特征的提取与分类。为了验证模型性能,采用凯斯西储大学轴承数据集进行轴承故障诊断分析,同时引入常见神经网络作为对比,检验不同模型的分类性能。结果表明,相较于其他图像编码方式与神经网络,该模型在载荷变化以及噪声污染时,仍保持了良好的诊断性能。
Aiming at the problem of relevant information between data being not able to be fully utilized during one-dimensional signals taken as input of convolutional neural network(CNN),a bearing fault diagnosis model based on GADF-CNN was proposed.Using Gram angle difference field(GADF)to encode the collected vibration signals,it was easy to perform angle perspective,identify the time correlation in different time intervals and produce the corresponding feature maps.Then,feature maps were input into the convolutional neural network(CNN)to adaptively complete extraction and classification of rolling bearing fault features.In order to verify the performance of the model,Case Western Reserve University,USA bearing data set was used to do bearing fault diagnosis analysis,and a common neural network was introduced to contrastively test the classification performance of different models.The results showed that compared with other image coding methods and neural networks,the proposed model still maintains good diagnostic performance during load variation and noise pollution.
作者
仝钰
庞新宇
魏子涵
TONG Yu;PANG Xinyu;WEI Zihan(College of Mechanical and Transport Engineering,Taiyuan University of Technology,Taiyuan 030024,China)
出处
《振动与冲击》
EI
CSCD
北大核心
2021年第5期247-253,260,共8页
Journal of Vibration and Shock
基金
国家自然科学基金(51805352)
山西省面上自然基金项目(201901D111062)。
关键词
轴承故障诊断
深度学习
格拉姆角差域
卷积神经网络
bearing fault diagnosis
deep learning
Gram angular difference field(GADF)
convolutional neural network(CNN)