An Embedded Estimator for Online Harmonic Monitoring in Power-Electronic Grids

A real-time harmonic estimator is proposed to provide a high-accuracy, computationally-efficient, and low-cache-cost solution that can be applied in embedded system for the online monitoring of numerous harmonics in wide frequency band in electronic power systems. Firstly, the fast-tuning filter bank, named FT-TFT FB, is presented, which can be fast generated by straightforwardly tuning the pseudo inverse of the basic single-frequency Taylor-Fourier basis matrix using the frequency-related diagonal weight matrix. Secondly, the robust FT-TFT estimator, named RFT-TFT, is proposed by estimating harmonics using the FT-TFT FB with a fast bias compensation process. It is verified using the numerical test that RFT-TFT estimator can achieve the same high accuracy as iterative TFT (I-TFT) estimator under steady and dynamic conditions but be much faster than I-TFT. The online application feasibility is further verified by implementing RFT-TFT estimator in C6748 DSP-based embedded prototype. Finally, this prototype is applied for the field signal analysis of
$\mathbf {\pm }800$
kV Ultra HVDC transmission system and the real-time harmonic monitoring of one 0.3 MW distributed photovoltaic (PV) system, respectively.