Frequency security constraint in unit commitment with detailed frequency response behavior of wind turbines

Abstract

With the high penetration of wind turbines, the necessity of incorporating frequency security considerations into power system scheduling rises. Existing methods achieve the explicit modeling of frequency security constraints by simplifying the frequency response behavior of wind turbines. However, simplifications might lead to inaccuracy. To address this issue, this paper models the frequency response from wind turbines in detail and proposes a novel framework to construct the frequency security constraint for unit commitment (UC). First, the frequency security constraint is positioned at the segment that is effective for the dispatch decision instead of the whole boundary, which is unnecessary and complicated. Then, an analytical linear surrogate expression of the frequency security boundary is constructed through a data-driven approach. To ensure the accuracy of the surrogate constraint, a neighborhood sampling strategy is proposed to collect balanced samples. Furthermore, to reduce the linearization error of the surrogate constraints, supplementary constraints are added to restrict the width of the surrogate constraint. Finally, to address the modeling errors that may deviate from the frequency security requirements, a correction strategy is proposed. Case studies validate the proposed method and verify that it exceeds existing methods in the modeling accuracy of the power system frequency security.