电力系统及其自动化学报2025，Vol.37Issue(7)：59-68,10.DOI:10.19635/j.cnki.csu-epsa.001542
基于分形特征的集合经验模态分解的谐波检测
Harmonic Detection Based on Fractal Ensemble Empirical Mode Decomposition
摘要
Abstract
Complementary ensemble empirical mode decomposition(CEEMD)is susceptible to noise interference in harmonic detection,resulting in false components and modal aliasing.To solve this problem,a harmonic detection method based on fractal ensemble empirical mode decomposition is proposed.First,the signal is decomposed using CEEMD to obtain a series of intrinsic mode functions(IMFs).Then,the fractal dimension is used to detect the random-ness of different IMFs.By selecting an appropriate box counting dimension threshold,abnormal components with exces-sively high box counting dimensions can be extracted,and the remaining signal is adaptively decomposed.Finally,the decomposed harmonic components are subjected to parameter detection using Hilbert transform to obtain the instanta-neous amplitude and frequency of each harmonic.The proposed method is used to detect the steady-state and time-vary-ing simulation signals,as well as the actual electric arc furnace signals,and it is also compared with other types of em-pirical mode decomposition algorithms and the fast Fourier transform method.Results show that this method has high noise robustness while ensuring a high detection accuracy,and it is not prone to false components.In addition,it can ef-fectively detect signals containing inter-harmonics.关键词
谐波检测/分形维数/模态混叠/希尔伯特变换Key words
harmonic detection/fractal dimension/modal aliasing/Hilbert transform(HT)分类
信息技术与安全科学引用本文复制引用
鲁亮,马建清..基于分形特征的集合经验模态分解的谐波检测[J].电力系统及其自动化学报,2025,37(7):59-68,10.基金项目
国家自然科学基金资助项目(61671338). （61671338）
