文章导读
总览 评价 王尊 1, , 顾祥柏 1,2, , 耿志强 1,* ( 1、 北京化工大学信息科学与技术学院,北京 100029; 2、 中国石化炼化工程(集团)股份有限公司,北京 100029; ) 摘要: 总体平均经验模态分解(EEMD)有效解决了传统希尔伯特-黄变换(HHT)中经验
王尊1,, 顾祥柏1,2,, 耿志强1,*
(
1、北京化工大学信息科学与技术学院,北京 100029; 2、中国石化炼化工程(集团)股份有限公司,北京 100029; )
摘要:
总体平均经验模态分解(EEMD)有效解决了传统希尔伯特-黄变换(HHT)中经验模态分解(EMD)模态混叠的问题,利用分解结果求得希尔伯特谱进行分析的方法,在故障诊断领域已经得到广泛应用。然而,该方法难以准确定位故障,不适于在线诊断。为此,提出了一种基于EEMD残差进行故障诊断的新方法,并使用基于移动窗格采样数据的EEMD,在线获得残差最大值的变化,实现了在线诊断故障并确定故障发生时间段。最后该方法在TE过程中实现,与传统的方法希尔伯特谱以及PCA诊断方法对比,验证了提出方法的有效性与优势。
关键词:
故障诊断;HHT;EEMD残差;TE过程
WANG Zun1,, GU Xiangbai1,, GENG Zhiqiang2,*
(
1、Information Science and Technology School, Beijing University of Chemical Technology, Beijing 10029; 2、Sinopec Engineering (Group) Co., LTD, Beijing 100101; )
Abstract:
Ensemble empirical mode decomposition (EEMD) has given an effective solution to the mode mixing problem of empirical mode decomposition (EMD) in traditional Hilbert-Huang Transform (HHT). The analysis method with Hilbert spectrum obtained by decomposition results has been widely used in the field of fault diagnosis. The method, however, is difficult to achieve accurate fault location and adapt to online diagnosis. Therefore, this paper proposes a new method based on EEMD residual. After processing the sampling data by EEMD according to moving windows, the variation of online obtained residual maximum can be used to diagnose faults online and identify fault time. Finally, the new method implemented in the TE process has a better application effect and advantage, comparing with the traditional HHT and PCA methods.
Tag:
点此返回栏目查看更多>>>参考论文