| 高宏， 王典， 张守京.基于CEEMD SE和LSTM的滚动轴承剩余寿命预测[J].轻工机械,2021,39(3):10-15 |
| 基于CEEMD SE和LSTM的滚动轴承剩余寿命预测 |
| Residual Life Prediction of Rolling Bearing Based on CEEMD SE and LSTM |
| |
| DOI：10.3969/j.issn.1005 2895.2021.03.002 |
| 中文关键词: 滚动轴承 剩余寿命预测 集成经验模态分解（CEEMD） 多频率尺度样本熵 长短期记忆神经网络 |
| 英文关键词:rolling bearing remaining useful life CEEMD multi frequency scale sample entropy LSTM （Long Short Term Memory） neural network |
| 基金项目:国家重点研发计划项目(2019YFB1707205)；西安市现代智能纺织装备重点实验室(2019220614SYS021CG043)；陕西省教育厅科研计划项目(17JK0321)。 |
|
| 摘要点击次数: 703 |
| 全文下载次数: 748 |
| 中文摘要: |
| 针对滚动轴承退化数据的复杂性和传统的寿命预测方法不能充分利用数据的相关性从而导致预测精度不高的问题，课题组提出了一种基于多频率尺度样本熵(SE)和长短期记忆神经网络(LSTM)相结合的寿命预测模型。该模型采用互补集成经验模态分解(CEEMD)结合相关系数分析，从滚动轴承振动信号中提取包含主要退化信息的IMF分量，并提取其样本熵矩阵，用于训练和测试LSTM。通过滚动轴承全寿命试验证明该模型可以准确预测滚动轴承剩余寿命，与BP神经网络和极限学习机(ELM)的预测效果对比验证了该模型的有效性。 |
| 英文摘要: |
| In view of the complexity of rolling bearing degradation data, and the low prediction accuracy caused by traditional life prediction methods cannot make full use of data correlation, a life prediction model based on the combination of Multifrequency scale sample entropy and LSTM （long short term memory） neural network was proposed. In this model, CEEMD (complementary ensemble empirical mode decomposition) combined with correlation coefficient analysis was used to extract IMF components containing major degradation information from rolling bearing vibration signals, and its sample entropy matrix was extracted for training and testing LSTM. Through the life test of rolling bearing, it is proved that the model can accurately predict the remaining life of rolling bearing, and the validity of the model is verified by comparing with the prediction results of BP neural network and ELM （extreme learning machine）. |
| 查看全文 查看/发表评论 下载PDF阅读器 |
| 关闭 |
