Swin Transformer Architecture-Based Power System Transient Stability Assessment

To further improve the classification performance of the data-driven intelligent assessment model for power system transient stability (TSA), this paper presents a Swin Transformer-based TSA of power system approach. Firstly, Swin Transformer employs a hierarchical model structure to boost feature extraction. Secondly, it alternately computes self-attention under fixed and shifted windows, allowing the model to extract local features related to transient stability more effectively and realize global modeling of the nonlinear mapping of transient stability. Simulation results on the IEEE 39-bus system demonstrate that the proposed method outperforms conventional transient stability assessment techniques in accuracy and improves the classification performance of the intelligent assessment model.