A fast and precise power system frequency estimation method without iterations

Abstract The conventional power system frequency estimation methods based on the Discrete Fourier Transform (DFT) normally require iterative calculations to eliminate the errors due to spectral leakage in off-nominal frequencies. It cannot provide both fast response and good accuracy under a system disturbance. In this paper, a fast and precise frequency estimation method is presented. The theoretical analysis shows the angle difference of the positive sequence voltage is exactly linear to the frequency deviation. Based on this angle characteristic, this method is able to precisely calculate the frequency deviation from angle differences between the continuous DFT calculations without the need for any iteration. Even if there is only one single phase voltage input available, the simulated positive sequence voltages can be derived by constructing the remaining two phase-voltages utilizing an appropriate phasor shift. To effectively restrain the harmonic interferences, a two-cycle Hanning window is applied as a filter. One method to detect the phase angle jump is proposed to avoid transient frequency errors by using the 2nd order derivative of the frequency. The simulation compares the frequency estimation performance between the proposed non-iterative method and the conventional iterative method. The simulation results show the frequency response is faster and the accuracy is reliable within the range 40–70 Hz, even if there are harmonic interferences or three-phase unbalances. This proposed frequency estimation method is easy to implement, and it can be widely applied in the intelligent electronic devices within the power system.