A Cyber-Physical-Aware MST-GNN for Power System Load Forecasting

Key Points

• Investigate an advanced model for forecasting power system loads using cyber-physical awareness and graph neural networks.
• Developed a MST-GNN model integrating cyber-physical aspects.
• Applied the model to historical power load data for analysis.
• Utilized advanced data analytics techniques to optimize forecasting accuracy.
• Demonstrated higher accuracy in load forecasting compared to traditional models.
• Identified significant improvements in predictive performance under various conditions.
• Provided insights into the impact of cyber-physical parameters on forecasting capabilities.

Abstract