DSP methods for dynamic estimation of frequency and magnitude parameters in power system transients

There are several developments in accurate dynamic estimation of the actual power line frequency-(APLF) and of the actual power-line harmonics (APLH) by digital filtering, which can be classified under 3 groups: (1) methods of phasor decomposition with local determination of rotation speed through relation of two orthogonal components; (2) methods of zero-crossing detection of power-line signal, accompanied with averaging, windowing and local (near zero-crossing points) interpolation; and (3) methods based on Fourier transform-fast and short-time (FFT, STFT) with interpolation. Different improvements of these methods are considered and compared here for achieving maximal accuracy concerning affecting factors-external (such as random noises, high harmonics, frequency deviation) and internal (such as resolution of A/D conversion and sampling frequency). Obtained simulation results together with considerations of their application for magnitude spectrum analysis are described.