Frequency-scaled optimized time-frequency transform for harmonic estimation in photovoltaic-based microgrid

Renewable (especially photovoltaic [PV])-based distributed generations (DGs) integrated via converter stations pose serious challenge against effective monitoring of fundamental and harmonic phasors because of the presence of time-varying Nth-order harmonics, which get escalated in terms of voltage flicker, total harmonics distortion profiles with increased penetration level of such renewable sources. The generalized S-transform (ST) has been used extensively for these nonstationary power signals analysis because it localizes the signal spectrum in time and yields a phase spectrum also for obtaining both amplitude and phase characteristics. However, the transform suffers from large execution time and arbitrary choice of window parameters. The paper, therefore, presents a new approach to extract time-frequency response (TFR) from a harmonically scaled Fourier kernel ST optimized by metaheuristic firefly algorithm suitable for monitoring of microgrid phasors in real time, which can be ultimately used for effective power management and coordination. This new formulation uses selective frequency scaling to achieve a significant time reduction (within a half cycle of the initiation of the disturbance) in monitoring and tracking time-varying harmonic disturbances, especially for DG-integrated operations. Extensive simulation results for monitoring of time-varying fundamental and harmonics in PV-based grid-connected power networks are presented against efficacy measures. Further, to validate in real time, a TMS320 C6713 digital signal processor (DSP) test bench is included in the paper for fundamental and harmonics estimation.