A fast parameter estimation method for high-frequency oscillation based on empirical wavelet transform and moving least square

Sci Rep. 2025 Jan 2;15(1):599. doi: 10.1038/s41598-024-84272-9.

Abstract

In renewable power systems, the interaction between generators, power electronic devices, and the grid has led to frequent high-frequency oscillation (HFO) events. These events can result in significant generation losses and pose serious threats to system stability. Therefore, the rapid and accurate HFO parameter estimation is crucial for early warning and effective mitigation of HFO. This paper proposes a fast estimation method for HFO parameters. The empirical wavelet transform (EWT) method is proposed to decompose HFO signals into individual oscillation mode within 1-fundamental-cycle window. The decomposed oscillation mode is then fitted to the actual oscillatory mode through the moving lease square (MLS) method. Subsequently, the frequency and magnitude of HFO are estimated based on the peaks and troughs of the fitted waveform data. Case studies demonstrate that the proposed EWT-MLS method achieves higher accuracy than other methods for HFO parameter estimation within 1-fundamental-cycle window. Additionally, the response time of the proposed method is shorter than 20 ms, highlighting its excellent rapid response capabilities.

Keywords: Empirical wavelet transform (EWT); High-frequency oscillation (HFO); Moving lease square (MLS); Parameter estimation.