IEEE

Constraints and Variables Reduction for Optimal Power Flow Using Hierarchical Graph Neural Networks with Virtual Node-Splitting

Abstract: Power system networks are often modeled as homogeneous graphs, which limits the ability of graph neural network (GNN) to capture individual generator features at the same nodes. By ...
IEEE

Remining Useful Life Prediction Using Graph Neural Networks for Rolling Bearing

Abstract: Remaining Useful Life (RUL) prediction is a necessary tool for condition monitoring and health management of rotating machinery, which is very important to ensure safe and economical ...