A novel frequency estimator for protection applications in active distribution networks

Abstract

Frequency is a useful parameter for ensuring the proper functioning of the power system operation, control, and protection. In this paper, a fast and accurate frequency estimator is developed using a Zero-crossing and a Peak, occurring in the buffered data instead of two Zero crossings, which is commonly used to measure the frequency. Prior to this, the signal is preprocessed using the Sliding Goertzel Filter (SGF) that mitigates the impact of undesirable features such as harmonics, noise, etc. from the signal. The two frequency estimators, with and without SGF, then work in parallel to form the hybrid estimator and named as ZC Pk − HFE. The estimated frequency is chosen based on the minimum error computed from the two parallel estimators. The performance of the estimator is, then, validated using the simulated signals polluted with harmonics, noise, transients etc. Moreover, the evaluation results with the IEEE-13 Active Distribution Network (ADN) incorporated with Photovoltaic DG using and dSPACE1104 and Real Time Digital Simulator (RTDS), show its capability to estimate real-time frequency with satisfactory output.