GAPSO-KELM在滚动轴承故障诊断中的应用
方俊豪，陈正坤，陈保家*，陈学力
（三峡大学 机械与动力学院，湖北 宜昌 443002）
摘要：为了减小故障特征提取对信号处理方法和人工经验的依赖性、降低诊断模型的计算复杂度、有效提高诊断精度，本文提出了一种卷积神经网络（CNN）与核极限学习机（KELM）相结合的滚动轴承故障诊断方法。首先，直接将不同故障模式下的滚动轴承原始振动信号进行分段处理，用以构建训练集、验证集和测试集。其次，利用CNN卷积运算提取特征，通过池化运算提炼简化特征。最后，将提取后的特征用来训练KELM，并采用遗传粒子群（GAPSO）算法对KELM的惩罚系数C与核参数σ进行优化设置，得到故障诊断模型。为评估方法有效性，采用同工况和变工况条件下进行实验测试，并与CNN-SVM、经典AlexNet 、VGG方法进行比较，结果显示该方法具有更好的准确性和稳定性。
关键词：卷积神经网络；滚动轴承；核极限学习机；故障诊断
中图分类号：TH133.33 文献标志码：A doi：10.3969/j.issn.1006-0316.2021.03.002
文章编号：1006－0316 (2021) 03－0009－08
Application of GAPSO-KELM in Rolling Bearing Fault Diagnosis
FANG Junhao，CHEN Zhengkun，CHEN Baojia，CHEN Xueli
( College of Mechanical and Power Engineering, China Three Gorges University, Yichang 443002, China )
Abstract：In order to reduce the dependence of fault feature extraction on signal processing methods and artificial experience, reduce the computational complexity of diagnosis model, and effectively improve the diagnosis accuracy, a rolling bearing fault diagnosis method combining convolutional neural networks (CNN) and kernel extreme learning machine (KELM) is proposed. Firstly, the original vibration signals of rolling bearing under different fault modes are directly processed by sections to construct training set, verification set and test set. Secondly, CNN convolution operation is used to extract the features and pool operation is used to extract the simplified feature information. Finally, the extracted features are used to train KELM, and the genetic particle swarm optimization (GAPSO) algorithm is used to set the penalty coefficient C and kernel parameter σ of KELM to obtain the fault diagnosis model. In order to evaluate the effectiveness of the method, experiments are carried out under the same and off-design working conditions, and compared with CNN-SVM, Alexnet and VGG methods, the results show that the method has better accuracy and stability.
Key words：convolutional neural network；rolling bearing；kernel extreme learning machine；fault diagnosis
———————————————
收稿日期：2020－10－09
基金项目：国家自然科学基金（51975324）；湖北省重点实验室开放基金（2020KJX02）；宜昌市应用基础研究项目（A19-302-08）
作者简介：方俊豪（1993－），男，湖北孝感人，硕士研究生，主要研究方向为机械故障诊断、深度学习。*通讯作者：陈保家（1977－），男，湖北孝感人，博士，教授，主要研究方向为机械装备状态监测、故障诊断及可靠性评估，E-mail：987794348@qq.com。
设为首页 | 加入收藏 | 免责条款
《机械》杂志版权所有 Copyright©2008-2012 Jixiezazhi.com All Rights Reserved
电话：028-85925070 传真：028-85925073 E-mail：jixie@vip.163.com
地址：四川省成都锦江工业开发区墨香路48号 邮编：610063
Powered by PageAdmin CMS