Enhanced Support Recovery for PMU Measurements Based on Taylor-Fourier Compressive Sensing Approach

Modern distribution networks are characterized by higher distortion and faster variability of voltages and currents. Accurate synchrophasor, frequency, and rate-of-change-of-frequency measurements thus ask for new techniques trying to reduce latency while limiting the impact of spurious components. In this respect, Taylor-Fourier multifrequency approach is a good candidate for phasor measurement units intended for distribution system applications. In this present article, we propose an enhanced version of this approach based on the joint application of window functions and iterative support refinement by means of the phasor first-order derivative. The performance of the algorithm is thoroughly characterized through extensive numerical simulation of nonstandard test conditions that reproduce the challenges of real-world scenarios, with fundamental dynamics superimposed on interfering tones. The reported results confirm the enhanced spectral support recovery, resulting in a remarkable improvement of estimation accuracy.