Cosine Windows in Interpolated DFT-based Method for an Accurate High-Frequency Distortion Assessment in Power Systems

The transformation of electrical networks in the context of the new smart grid paradigm unavoidably involves new challenges regarding Power Quality (PQ) disturbances not only for customers but also for all the other involved stakeholders. Among PQ disturbances, waveform distortions have recently gained growing interest due to the massive presence of new technologies in distributed energy resources, in modern loads and in advanced smart metering systems. The presence of these devices determines arduous electromagnetic compatibility problems since the current and voltage waveform distortions in smart grids are characterized by spectral components above the traditional 2 kHz frequency limit, in a range extended up to 150 kHz. In this paper, an interpolated DFT-based (IDFT) method, recently proposed in the relevant literature in the field of signal processing, is properly extended for an accurate and fast assessment of power system waveform distortions in the frequency range from 2 to 150 kHz. Since DFT-based methods can suffer well-known spectral leakage problems, in this paper the IDFT is applied using cosine windows that minimize interference conditions among spectral components and maximise the estimation accuracy of the spectral component amplitude, phase angle and frequency. An optimal number of cosine window terms is also searched to improve the spectral analysis of high-frequency power system waveforms. Numerical applications on synthetic test signals and measured waveforms are carried out to quantify the accuracy and computational efforts of the proposed approach and to select the cosine window terms that better optimize the waveform distortion assessment.