变分模态分解在自动机故障诊断中的应用  

Application of variational mode decomposition in fault diagnosis of automaton

作　　者：安邦[1] 潘宏侠[1,2] 张玉学[1] 赵雄鹏[1] 

机构地区：[1]中北大学机械与动力工程学院,山西太原030051 [2]中北大学系统辨识与诊断技术研究所,山西太原030051

出　　处：《中国测试》2017年第7期112-116,共5页China Measurement & Test

基　　金：国家自然科学基金项目(51175480;51675491)

摘　　要：由于自动机工作环境复杂、各种响应信号相互叠加,为准确、高效地提取自动机信号的故障特征,提出一种应用变分模态分解(VMD)和极限学习机(ELM)的自动机故障诊断方法。首先对自动机信号进行变分模态分解,并与经验模态分解(EMD)结果进行对比;同时提取各模态分量的能量百分比和各工况下不同样本的样本熵作为特征值;将提取到的特征值输入到极限学习机中进行故障诊断,再与传统的双谱分析诊断结果进行比较。最终VMD方法实现信号频域内各分量的自适应剖分,并得出ELM的故障诊断准确率为87.5%。实验结果表明:变分模态分解能够有效避免模态混叠现象,同时验证所提方法的可行性与有效性。Due to the complex working environment of automatons and the superimposition of various corresponding signals, to extract the fault characteristics of the signal accurately and efficiently, an automaton fault diagnosis method based on variational mode decomposition （VMD） and extreme learning machine （ELM） is proposed. Firstly, the VMD of automaton signal is performed and compared with empirical mode decomposition （EMD） results. Meanwhile, the energy percentage of each component and the sample entropy of each sample are extracted and taken as the eigenvalues. Then, the extracted feature parameters are input to the extreme learning machine （ELM） for fault diagnosis, and compared with the traditional bispectrum diagnostic results. Finally, the VMD method achieved the adaptive subdivision of the components in the signal frequency domain, and the accuracy of the ELM is 87.5%. The results showed that the VMD can effectively avoid the phenomenon of modal mixture, and verified the feasibility and effectiveness of the proposed method.

关 键 词：变分模态分解 模态混叠 极限学习机 自动机 故障诊断 

分 类 号：TP277[自动化与计算机技术—检测技术与自动化装置]