MoE-GraphSAGE-Based Integrated Evaluation of Transient Rotor Angle and Voltage Stability in Power Systems

TL;DR

This work tackles the challenge of online, accurate, and efficient joint assessment of transient rotor angle stability (TAS) and transient voltage stability (TVS) in modern power systems. It introduces MoE-GraphSAGE, a graph neural network framework that combines GraphSAGE for spatiotemporal grid features with a gated mixture-of-experts to handle diverse instability modes in a single multi-task model. The approach enables concurrent TAS/TVS classification and margin regression, validated on an improved IEEE 39-bus system where it achieves high accuracy (≈98.7%) and strong margin prediction (low MSE/MAE). The model runs in real time on standard accelerators and outperforms several baselines in both accuracy and efficiency, offering a practical solution for online stability assessment in increasingly complex power networks.

Abstract

The large-scale integration of renewable energy and power electronic devices has increased the complexity of power system stability, making transient stability assessment more challenging. Conventional methods are limited in both accuracy and computational efficiency. To address these challenges, this paper proposes MoE-GraphSAGE, a graph neural network framework based on the MoE for unified TAS and TVS assessment. The framework leverages GraphSAGE to capture the power grid's spatiotemporal topological features and employs multi-expert networks with a gating mechanism to model distinct instability modes jointly. Experimental results on the IEEE 39-bus system demonstrate that MoE-GraphSAGE achieves superior accuracy and efficiency, offering an effective solution for online multi-task transient stability assessment in complex power systems.

Figures (3)

• Figure 1: Architecture of the MoE-GraphSAGE Model
• Figure 2: TVS&TAS integrated assessment framework
• Figure 3: Computational efficiency of different algorithms