A Modified DFT-Based Phasor Estimation Algorithm Using an FIR Notch Filter

Accurate and fast extraction of the fundamental parameters of fault signals is a critical issue in the overall performance of protective relays. The decaying DC terms (DDTs) in the fault current challenge the discrete Fourier transform (DFT), the reference algorithm for phasor estimation. To meet this challenge, a new modification for half and full-cycle DFT algorithms is proposed in this paper. In the first step, a finite impulse response (FIR) notch filter designed for fundamental frequency is applied to the output of the DFT algorithm. After that, the outcome of the FIR notch filter and its complex conjugate make it possible to extract error terms. Finally, the estimated phasors are emptied from DDTs error. The ability of the proposed method for estimating phasor and tolerating disturbances was assessed using practical simulated signals. The results perform that the proposed method is much more robust, accurate, and faster than previous methods and enhances the efficiency of the DFT algorithm faced by multiple DDTs.

I. Introduction

The discrete Fourier transform (DFT) as a reference algorithm is widely used to estimate the fundamental phasor in protection applications of the power systems. The precise and rapid phasor estimation is essential to the proper decision-making and operation of digital relays, and the DFT algorithm meets this task. However, the security of the DFT algorithm is vulnerable to decaying DC terms (DDTs), and the corresponding estimated phasor severely oscillates until the DDTs decay. The DDT is a non-periodic signal created in the fault current owing to the ohmic-inductive feature of the power system. Basically, the creation of high error in current phasor estimation (up to 20% [1]) due to DDTs is a vital problem in DFT-based phasor estimation that several methods have been proposed to treat. These methods have been categorized and described in this paper as follows: