Frequency Constrained Scheduling in Renewable-Dominated Power Systems via Data-Driven Security Rule

The low inertia and constrained reserves in renewabledominated power systems present considerable challenges to frequency security. To mitigate the risk of load-shedding caused by low-frequency events following power imbalances, frequency constraints have increasingly been incorporated into scheduling frameworks in recent years. This study introduces a hybrid machine learning approach to extract the complex frequency nadir constraint from frequency dynamics and reformulate it into a convex and concise representation. The proposed method overcomes the limitations of previous approaches, which did not simultaneously account for the combined effects of diverse resources’ primary frequency response reserves, system inertia, and load frequency relief on the frequency nadir constraint. This advancement is essential for enhancing the economic efficiency of system operation. Case studies on two test systems demonstrate the superiority of the proposed method compared to existing approaches. Lastly, potential scheduling applications of this method are explored.