现代电子技术2024，Vol.47Issue(24)：120-130,11.DOI:10.16652/j.issn.1004-373x.2024.24.019
基于深度学习的滚动轴承故障预测
Rolling bearing fault prediction based on deep learning
摘要
Abstract
In order to solve the problem of unsupervised fault detection and fault prediction of full-life rolling bearings with performance degradation and faults,a method of bearing fault detection and prediction based on deep learning is proposed.The vibration signals of the rolling bearing are divided into different stages,the Wasserstein generative adversarial network gradient penalty(WGAN-GP)is used to calculate monitoring statistics of vibration signals,and the control limits are established to detect the vibration signals.The wavelet packet threshold denoising-successive variational mode decomposition(WPTD-SVMD)is used conduct the denosing process of the abnormal part of the rolling bearing vibration signal,and convolutional neural networks-bidirectional long/short term memory network(CNN-BiLSTM)is used to predict the fault degradation trend of the pre-processed signal.The effectiveness of the proposed method is verified by means of NASA lifetime bearing fault dataset.The experimental results verify that the proposed method can accurately distinguish between normal and faulty data in the fault detection phase,and can effectively retain essential information regarding fault characteristics and eliminate the impact of noise in the fault prediction stage,which has a relatively accurate prediction of the trend of fault amplitude.关键词
滚动轴承/生成对抗网络/卷积双向长短期记忆网络/故障检测/故障预测/深度学习Key words
rolling bearing/generate adversarial network/convolutional neural networks-bidirectional long/short term memory network/fault detection/fault prediction/deep learning分类
信息技术与安全科学引用本文复制引用
张晋恺,马洁..基于深度学习的滚动轴承故障预测[J].现代电子技术,2024,47(24):120-130,11.基金项目
国家重点研发计划子课题(2019YFB1705403) （2019YFB1705403）
国家自然科学基金面上项目(61973041) （61973041）
