基于TET与DSRNet-AttBiLSTM的滚动轴承剩余使用寿命预测

被引：0

作者：

周玉国 ^{[1
]}

张金超 ^{[1
]}

孙伊萍 ^{[1
]}

于春风 ^{[2
]}

周立俭 ^{[1
]}

机构：

[1] 青岛理工大学信息与控制工程学院

[2] 海军航空大学青岛校区航空仪电控工程与指挥系

来源：

振动与冲击 | 2024年 / 43卷 / 19期

关键词：

滚动轴承; 剩余使用寿命(RUL); 注意力机制; 特征提取;

D O I：

暂无

中图分类号：

TP183 [人工神经网络与计算]; TH133.33 [滚动轴承];

学科分类号：

摘要：

为了在滚动轴承剩余使用寿命(remaining useful life, RUL)预测中，能够准确地提取轴承的退化特征并进行有效的RUL预测。提出一种基于暂态提取变换(transient extracting transform, TET)与DSRNet-AttBiLSTM的滚动轴承RUL预测方法。首先，对原始振动信号分段重组后进行TET得到时频图，使用双线性插值对时频图进行降维，将降维后的时频图进行通道拼接得到轴承的时频图像化特征。其次，为了准确且有效地提取滚动轴承的退化特征，构建了包含深度可分离卷积和空间通道注意力的SConv和DConv基础模块，以此为基础建立了DSRNet来提取空间与通道两个维度下的轴承退化特征。再次，为了使双向长短时间记忆(bidirectional long short-term memory, BiLSTM)网络在学习时更加关注具有更重要信息的输入特征，在特征输入端构建了注意力层，并与BiLSTM相结合组成AttBiLSTM预测模块进行HI的计算。最后，使用线性回归拟合来预测滚动轴承的RUL。在PHM2012数据集与XJTU-SY数据集上试验的结果表明此方法能有效预测滚动轴承的RUL。

引用

页码：163 / 173

页数：11

共 30 条

[1] Generalized multiscale feature extraction for remaining useful life prediction of bearings with generative adversarial networks.[J].Suh Sungho;Lukowicz Paul;Lee Yong Oh.Knowledge-Based Systems.2022,
[2] Deep transfer learning based on dynamic domain adaptation for remaining useful life prediction under different working conditions.[J].Cheng; Han;Kong; Xianguang;Wang; Qibin;Ma; Hongbo;Yang; Shengkang;Chen; Gaige.Journal of Intelligent Manufacturing.2021, 2
[3] Spatiotemporal non-negative projected convolutional network with bidirectional NMF and 3DCNN for remaining useful life estimation of bearings.[J].Wang Xu;Wang Tianyang;Ming Anbo;Zhang Wei;Li Aihua;Chu Fulei.Neurocomputing.2021,
[4] Rolling Bearing Fault Diagnosis using Transient-extracting Transform and Linear Discriminant Analysis.[J].Zhou Yuguo;Yan Shaoting;Ren Yanbo;Liu Shiliang.Measurement.2021, prepublish
[5] A Probabilistic Framework for Remaining Useful Life Prediction of Bearings.[J].Wang Teng;Liu Zheng;Mrad Nezih.IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT.2021,
[6] An enhanced encoder–decoder framework for bearing remaining useful life prediction.[J].Liu Lu;Song Xiao;Chen Kai;Hou Baocun;Chai Xudong;Ning Huansheng.Measurement.2020, prepublish
[7] Supervised Health Stage Prediction Using Convolutional Neural Networks for Bearing Wear
Suh, Sungho
Jang, Joel
Won, Seungjae
Jha, Mayank Shekhar
Lee, Yong Oh
[J]. SENSORS, 2020, 20 (20) : 1 - 19
[8] Remaining useful life prediction for machinery by establishing scaled-corrected health indicators.[J].Hanbo Yang;Zheng Sun;Gedong Jiang;Fei Zhao;Xuesong Mei.Measurement.2020, prepublish
[9] Data-driven remaining useful life prediction via multiple sensor signals and deep long short-term memory neural network.[J].Jun Wu;Kui Hu;Yiwei Cheng;Haiping Zhu;Xinyu Shao;Yuanhang Wang.ISA Transactions.2020,
[10] A novel deep learning method based on attention mechanism for bearing remaining useful life prediction.[J].Yuanhang Chen;Gaoliang Peng;Zhiyu Zhu;Sijue Li.Applied Soft Computing Journal.2020,

← 1 2 3 →