Data-Driven Frequency Stability Assessment of Power Systems Considering Wind Power Uncertainty

Due to the high penetration of renewable energy sources (RESs), the inertia levels of power systems decrease, causing the increasing risks of frequency instability. In this paper, a data-driven frequency stability assessment (FSA) method with scenario-based wind uncertainty modeling is proposed. First, a scenario-based procedure for wind uncertainty modeling is proposed. Moreover, the leading physical factors of FSA problems are analyzed and extracted as the basic input features. The input feature enhancement is established by solving rotor motion equations. Then, a physics-guided gated recurrent unit (PG-GRU) based FSA method is introduced. Finally, through case studies, the accuracy, rapidity, and applicability of the proposed FSA method under small data sets are validated in IEEE 10-machine 39-bus power system, and the advantages of adopting scenario-based wind power modeling are analyzed.