Investigation on the burst noise detection of inverter based on the dynamic multi-distinguish analysis empirical mode decomposition

Autor: Jie Wu, Shen Yaru, Chen Dongyang, Xiaojuan Chen
Rok vydání: 2015
Předmět:
Zdroj: 2015 International Conference on Noise and Fluctuations (ICNF).
DOI: 10.1109/icnf.2015.7288561
Popis: In order to detect the burst noise of inverter accurately, a novel method is presented based on dynamic multi-distinguish analysis empirical mode decomposition (D-MEMD). The method is introduced the technique of multi-distinguish analysis to modify EMD, then the obtained IMFs are denoised by using dynamic adaptive threshold strategy, and finally instantaneous amplitude, instantaneous frequency and instantaneous phase are gained by applying the Hilbert transform into IMFs. The experiment results indicate that the method conquers mode aliasing successfully, avoids the occurrence of fault mode effectively, and can achieve the burst noise of inverter detection accurately.
Databáze: OpenAIRE