Static Voltage Stability Margin Prediction Based on LightGBM

In the new power system, renewable energy source is continuously integrated into the power grid. Due to the randomness and uncertainty of renewable energy source, the probability of random fluctuations in the direction of new energy generation or load continues to increase. Static voltage stability analysis in a fixed direction can no longer meet the prediction requirements of static voltage stability margin (SVSM) in the new power systems. A SVSM prediction method based on light gradient boosting machine (LightGBM) is proposed, which comprehensively considers the initial operation point and the power growth direction. The simulation analysis based on IEEE 39 node system verified the effectiveness of the LightGBM model in SVSM prediction.