摘要
针对滚动轴承早期故障信号较弱及特征数据提取效果差,导致故障诊断准确率低以及故障诊断效率低的问题,提出一种信号处理结合深度神经网络的故障诊断方法。首先,采用变分模态分解(VMD)法提取主轴承振动数据中的特征数据;然后为了确定VMD算法中最佳的模态分量个数K及惩罚参数α,增强特征提取的效果,将最小排列熵作为适应度函数,采用全局优化能力强的正弦混沌自适应鲸鱼优化算法(CAWOA)进行参数的确定,得到最优模态分量;接着,根据最优模态分量构造特征向量,将特征向量作为CNN-BiLSTM网络的输入,实现故障的分类。最后,根据实验平台采集的数据进行实验分析。结果表明,优化VMD-CNN-BiLSTM轴承故障诊断模型相较于其他故障诊断模型,在准确率以及实时性上均有明显提升。
In allusion to the problems of low fault diagnosis accuracy and low fault diagnosis efficiency caused by weak early fault signal and poor feature data extraction effect of rolling bearing,a fault diagnosis method combining signal processing and deep neural network is proposed.The variational mode decomposition(VMD)is used to extract the feature data of main bearing vibration data.In order to determine the optimal number of modal components K and penalty parameters in the VMD algorithmα,and enhance the effectiveness of feature extraction,the minimum permutation entropy is used as the fitness function,and the sine chaos adaptive whale optimization algorithm(CAWOA)with strong global optimization ability is used to determine the parameters and obtain the optimal modal component.The feature vector is constructed based on the optimal modal components,which is used as inputs of the CNN(convolutional neural network)BiLSTM(bidirectional long short term memory)network to realize the fault classification.Based on the data collected from the experimental platform,the experimental analysis results show that in comparison with other fault diagnosis models,the optimized VMD-CNN-BiLSTM bearing fault diagnosis model can significantly improve accuracy and real-time performance.
作者
曹景胜
于洋
王琦
董翼宁
CAO Jingsheng;YU Yang;WANG Qi;DONG Yining(School of Information Science and Engineering,Shenyang University of Technology,Shenyang 110870,China;College of Automobile and Traffic Engineering,Liaoning University of Technology,Liaoning University of Technology,Jinzhou 121001,China)
出处
《现代电子技术》
北大核心
2024年第12期115-121,共7页
Modern Electronics Technique
基金
国家自然科学基金项目(51675257)
国家自然科学基金青年基金项目(51305190)
辽宁省教育厅基本科研项目(面上项目)(LJKMZ20220976)
辽宁省自然科学基金指导计划项目(20180550020)。
