Shifting window average method for accurate frequency measurement in power systems

A shifting window average method (SWAM) is proposed to enhance the accuracy of the widely implemented discrete Fourier transform (DFT)-based algorithm in power system frequency measurements. First, the authors have comprehensively analyzed the error induced in the DFT-based algorithm. The SWAM is then introduced to improve frequency measurement accuracy by examining the analytic form of the error. Test results have validated that it achieves higher accuracy compared to the DFT-based algorithm. Due to its high accuracy and low required processing effort, the SWAM is a valuable candidate for online frequency estimation in real-life power systems, especially when harmonics, uncharacteristic harmonics, noises, and large frequency excursions exist.