Denoising of Electric Power System Signals by ICA Based on EMD Virtual Channel

The method integrated by independent component analysis (ICA) and empirical mode decomposition (EMD) is proposed in this paper, which can separate the signal from white noise. We used this method to eliminate the white noise from electric power system signals. The detailed courses of construction of virtual noise channel used EMD are described. Based on the characteristic of signals and noise, the principle of select intrinsic mode function (IMF), using the Hilbert time- frequency spectra, is proposed to construct virtual noise channel. Then the inputs of the ICA are reconstructed by virtual channel and observation signal. By the ICA, the signal-noise separation will be realized. The simulation tests indicate that the de-noising performance of electric power system signals is effective.