文章基本信息

标题：Denoising of Industrial Oscillation Data Using EEMD with CCA ⁎
本地全文：下载
作者：Xun Lang ; Yan Liu ; Yufeng Zhang 等
期刊名称：IFAC PapersOnLine
印刷版ISSN：2405-8963
出版年度：2020
卷号：53
期号：2
页码：11662-11668
DOI：10.1016/j.ifacol.2020.12.655
语种：English
出版社：Elsevier
摘要：AbstractIndustrial oscillation recordings are often contaminated with random noise, process disturbances and underlying nonstationarity, which obscure the useful information of the signal and complicate subsequent oscillation detection and diagnosis. This paper proposes a novel denoising technique to improve the quality of oscillation data, by jointly employing ensemble empirical mode decomposition (EEMD) with canonical correlation analysis (CCA). The proposed method first utilizes EEMD to decompose the single-loop data into a set of intrinsic mode functions (IMFs). Then CCA is applied to isolate the oscillation-dependent components from the decomposed IMFs. We evaluated the performance of the method through both numerical and industrial examples. The results demonstrate that this work is a promising tool for oscillation data preprocessing in the single control-loops.
关键词：KeywordsIndustrial OscillationDenoisingEEMDCCA