Open AccessFeature Extraction of Conducted Electromagnetic Noise Based on KPCA and Its Application in Main Network
Author(s) -
Min Lü,
Junshuo Huang,
Minghui Sun
Publication year - 2019
Publication title -
iop conference series. materials science and engineering
Language(s) - English
Resource type - Journals
eISSN - 1757-899X
pISSN - 1757-8981
DOI - 10.1088/1757-899x/631/4/042052
Subject(s) - noise (video) , signal (programming language) , pattern recognition (psychology) , interference (communication) , time domain , principal component analysis , artificial intelligence , feature extraction , computer science , frequency domain , feature (linguistics) , transformation (genetics) , kernel principal component analysis , identification (biology) , feature vector , support vector machine , telecommunications , computer vision , kernel method , channel (broadcasting) , linguistics , philosophy , biochemistry , chemistry , botany , biology , image (mathematics) , gene , programming language
In this paper, a new method for conducting EMI noise source identification is proposed. Firstly, the basic principle of the classical KPCA method is analyzed. Then, the core principal component (KPCA) data is selected, and the input space is transformed into the feature space through nonlinear transformation. This method analyzes the relationship between the time and frequency of the electromagnetic interference signal from the frequency domain, so as to extract the time-domain characteristics of the noise signal, and finally diagnose the characteristics that cause the conduction of excessive signal, and according to the characteristics of the noise signal targeted suppression experiment.