Narrowband Interferers Cancellation Versus Estimation: Impact on Synchrophasor Measurement

One of the open issues in Phasor Measurement Units (PMU) design is how to suppress the influence of harmonics and Out-of-Band Interharmonics (OOBIs) on synchrophasor measurement over short observation intervals. In most PMUs, such narrowband interferers are removed before or during synchrophasor estimation. However, several researchers noticed that estimating harmonics and/or OOBIs is beneficial not only to extract more information from the acquired signal, but also to indirectly improve the estimation accuracy of the parameters of the fundamental component. At the moment, the pros and cons of these two opposite approaches are not fully clear, mainly due to the difficulty to draw general and fair conclusions from the comparison of different algorithms. In this paper, we partially address this problem by comparing two algorithms that rely on both a common theoretical background and the same final estimation stage, i.e., the Whitening-based Tuned Interpolated Dynamic Discrete Fourier Transform (WTpD2FT) and the Enhanced Interpolated Dynamic Discrete Fourier Transform (eIpD2FT). The former algorithm attempts to transform narrowband disturbances into white noise, whereas the latter detects and estimates the parameters just of the most significant interferers. Several simulation results suggest that the estimation of the narrowband interferers' parameters generally returns high-accuracy values of synchrophasor, fundamental frequency and rate of change of frequency (ROCOF) especially when large OOBIs affect the fundamental component, whereas the whitening-based disturbance cancellation approach is slightly preferable in the presence of multiple harmonics and growing noise levels.