New modified DFT-Prony-based algorithms for removal of decaying DC components from fundamental phasor estimates

Abstract

The paper proposes three new algorithms for estimation of the fundamental phasor in a power system, based on the removal of exponentially decaying DC components (DDCs). These components, as well as high-order harmonics and noise components, have a considerable effect on the accuracy and speed of convergence in numerical and digital relays speed of the protection relay operation. A discrete Fourier transform based approach with a modified Prony method was used to calculate and remove the unwanted effect of DDCs in a time interval slightly longer than the period of the fundamental harmonics. The use of an FIR notch filter leads to an algorithm that offers the possibility to estimate the parameters of unwanted DDCs in a simpler and analytically more precise way, thus facilitating their program implementation. The first two algorithms offer the ability to easily adjust the response speed detection time. This flexibility of the algorithms provides a compromise in terms of response speed and expected reliability and security of fault detection. In the third algorithm developed based on the application of the Least Error Squares method, new analytical expressions for estimating the parameters of the processed current signal are derived. The developed algorithms have low numerical and computational complexity, while maintaining their high performances even in conditions of a very strong noise signal. The simulation results for different test signals demonstrate high precision in the estimation of fundamental phasor of the proposed algorithms.