Design of Compressive Sensing Adaptive Taylor-Fourier Comb Filters for Harmonic Synchrophasor Estimation

In modern power systems, phasor measurements are expected to deal with challenging conditions, e.g., fast dynamics and high distortion levels. Taylor-Fourier Multifrequency models represent a promising solution, but their performance is strongly related to the accurate extraction of the signal spectral support. In this context, this paper proposes an enhanced method for support recovery that exploits the inherent block-sparsity properties of electrical signals. The proposed method is fully characterized in diverse and distorted test conditions, inspired by reference standards and real-world scenarios. The comparison against another Compressive Sensing based approach confirms the significant improvement in terms of both recovered support exactness and synchrophasor measurement accuracy.