Detection of voltage variation events using variational mode decomposition

Abstract

This paper presents a method of detecting voltage sag, swell and interruption events based on a signal processing tool viz. variational mode decomposition (VMD). The proposed method is able to keep track of variations in amplitude, phase, frequency and undelivered energy due to such events. Voltage signal is half-wave rectified and then decomposed by VMD to get DC part representing instantaneous amplitude envelope, thereby obtaining start/end times and depth of sag/swell. The phase jump and frequency deviations are observed from instantaneous phase, obtained from analytic signal of VMD mode having center frequency nearer to fundamental. Algorithm is tested for various cases of synthetic and simulated signals in MATLAB/Simulink. Comparative performance with recursive discrete Fourier transform and Hilbert transform based method is demonstrated.