Small signal stability prediction and correction control algorithm for wind power systems based on LightGBM

With the increasing penetration of wind power in electrical power systems, its impact on small signal stability has become increasingly significant. The volatility and uncertainty introduced by wind power integration pose new challenges to power system stability analysis and control. To accurately predict and correct the small signal stability of wind power systems, this paper proposes a prediction and correction method based on the Light Gradient Boosting Machine (LightGBM) algorithm. The LightGBM model offers high computational efficiency and low memory consumption, enabling large-scale data processing and automatic handling of missing values, thereby accurately extracting grid characteristics. To verify the reliability of the proposed method, Gaussian white noise with three different signal-to-noise ratios is introduced to evaluate the model’s performance and robustness. Case studies are conducted using a 3-machine 9-bus system and a 10-machine 39-bus system, in which certain conventional generators are replaced with aggregated wind farms. By applying the LightGBM model to predict the system’s minimum damping ratio and constructing a correction optimization model based on damping ratio sensitivity, unstable operating states are effectively adjusted. Simulation results demonstrate that the proposed method achieves high prediction accuracy and rapid correction response, confirming its feasibility and effectiveness in small signal stability analysis and control of wind-integrated power systems.