Abstract and subjects
State estimation is highly critical for accurately observing the dynamic behavior of the power grids and minimizing risks from cyber threats. However, existing state estimation methods encounter challenges in accurately capturing power system dynamics, primarily because of limitations in encoding the grid topology and sparse measurements. This paper proposes a physics-informed graphical learning state estimation method to address these limitations by leveraging both domain physical knowledge and a graph neural network (GNN). We employ a GNN architecture that can handle the graph-structured data of power systems more effectively than traditional data-driven methods. The physics-based knowledge is constructed from the branch current formulation, making the approach adaptable to both transmission and distribution systems. The validation results of three IEEE test systems show that the proposed method can achieve lower mean square error more than 20% than the conventional methods.
•We propose the physics-informed graphical learning approach, combining model-based dynamic state estimation and spatial-temporal graph neural network.•Model-based state estimation uses high-accuracy Kalman filter for μPMU measurements. The graph neural network yields estimated states with both μPMU measurements and traditional measurements.•The graph neural network considers the data under both temporal and spatial correlations, enhancing the training process through its combination with model-based Kalman filtering since model-based estimation contains physical information of the system.•Comprehensive case studies are performed with IEEE 5-bus, 123-bus and 8500-bus systems.