Taylor–Fourier Multifrequency Approach to Power Quality Monitoring in DC Grids

Direct current (dc) grids and microgrids represent attractive solutions for the integration of renewable energy sources, storage systems, and chargers for electric vehicles, all requiring power electronic interface converters. However, there is a severe lack of dedicated measurement tools, both from a scientific and normative point of view. The present article proposes to extend the concept of wide-area measurements synchronized with respect to an absolute time reference, like the coordinated universal time (UTC) timescale, to dc distribution, which enables observing the state of the grid, tracking dynamics, and performing power quality (PQ) monitoring, in particular measuring the most significant ripple components. In this respect, a class of algorithms, based on a suitably modified Taylor Fourier multifrequency (TFM) model, has been developed. Two methods for retrieving the frequency components to be included in the TFM model have been compared: a compressive-sensing-based algorithm exploiting the spectral sparsity of the signal and a new approach using the estimation of signal parameters by rotational invariance technique (ESPRIT) and clustering of spectral components. Achieved performance is assessed through numerical simulations that mimic a realistic situation. Results highlight the potentialities of the proposed approaches, which provide highly accurate estimates of the most significant spectral components superimposed to the dc voltage, both under steady-state and dynamic conditions.