Machine Learning Based Method for Transient Stability Assessment of Power Systems

Machine learning is playing an increasingly important role in transient stability assessment. However, as the scale of power grids gradually expands, the number of data dimensions also increases greatly, which places higher demands on the performance of the model. In this paper, based on the physical characteristics of the power grid, a feature set that can reflect the operating state of the power system is defined and a Pearson correlation analysis is performed. The XGBoost model is then constructed and applied to the transient stability assessment process, which solves the shortcomings of traditional machine learning models in terms of poor interpretation and draws decision boundaries for different classification models for the important features. The results show that the XGBoost algorithm has good predictive performance in terms of accuracy and recall.