Data-driven Stability Assessment for New Power Systems: Typical Characteristics, Research Framework and Future Prospects

The widely integrated renewable energy sources bring the new power systems (NPS) far more nonlinearity, flexibility, randomness and openness than the conventional one. The stability mechanism of the NPS is profoundly changed with the stable operation facing serious challenges. Hence, accurate stability assessment (SA) of NPS is essential to ensure the stable and reliable operation. Data-driven SA method is gaining more and more attention because of its superior performance on massive data processing, nonlinear relationship characterization and accurate assessment. Accordingly, this paper first summarizes the characteristics of NPSs, then discusses the connotation, characteristics, trends and framework of SA suitable for NPSs. The recent data-driven SA researches are analyzed from two dimensions: conventional data-driven and deep learning-driven, and further classified according to the data characteristics from three perspectives: discrete, time-series and spatial-temporal data. The applicability of key technologies and their advantages in the NPS is reviewed and evaluated. Finally, the research gaps and future trends are summarized and outlooked from the characteristics of the NPS.