Prediction and mitigation of nonlocal cascading failures using graph neural networks

Abstract

Cascading failures (CFs) in electrical power grids propagate nonlocally; After a local disturbance, the second failure may be distant. To study the avalanche dynamics and mitigation strategy of nonlocal CFs, numerical simulation is necessary; however, computational complexity is high. Here, we first propose an avalanche centrality (AC) of each node, a measure related to avalanche size, based on the Motter and Lai model. Second, we train a graph neural network (GNN) with the AC in small networks. Next, the trained GNN predicts the AC ranking in much larger networks and real-world electrical grids. This result can be used effectively for avalanche mitigation. The framework we develop can be implemented in other complex processes that are computationally costly to simulate in large networks.

References in corpus (35)

• A Comprehensive Survey on Graph Neural Networks
• Catastrophic cascade of failures in interdependent networks
• Epidemic processes in complex networks
• Cascade-based attacks on complex networks
• Efficient Immunization Strategies for Computer Networks and Populations
• A model for cascading failures in complex networks
• Cascade control and defense in complex networks
• Modeling Cascading Failures in the North American Power Grid
• k-core organization of complex networks
• Discrete-time Markov chain approach to contact-based disease spreading in complex networks
• Avalanche Collapse of Interdependent Network
• Dynamically induced cascading failures in power grids
• Avoiding catastrophic failure in correlated networks of networks
• Sandpile on Scale-Free Networks
• Percolation in Multiplex Networks with Overlap
• Vertex overload breakdown in evolving networks
• k-core (bootstrap) percolation on complex networks: Critical phenomena and nonlocal effects
• Cascading Power Outages Propagate Locally in an Influence Graph that is not the Actual Grid Topology
• Abrupt phase transition of epidemic spreading in simplicial complexes
• Containing epidemic outbreaks by message-passing techniques
• Immunization and targeted destruction of networks using explosive percolation
• A Random Growth Model for Power Grids and Other Spatially Embedded Infrastructure Networks
• Effective approach to epidemic containment using link equations in complex networks
• Non-perturbative heterogeneous mean-field approach to epidemic spreading in complex networks
• A High Robustness and Low Cost Model for Cascading Failures
• Hybrid Phase Transition into an Absorbing State: Percolation and Avalanches
• Robustness of the avalanche dynamics in data packet transport on scale-free networks
• Critical dynamics of the k-core pruning process
• Deciphering the imprint of topology on nonlinear dynamical network stability
• Effective epidemic containment strategy in hypergraphs
• Critical behavior of $k$-core percolation: Numerical studies
• Probabilistic prediction in scale-free networks: Diameter changes
• Emergent rewirings for cascades on correlated networks
• Distribution of blackouts in the power grid and the Motter and Lai model
• Effective vaccination strategy using graph neural network ansatz

Cited by in corpus (6)

• Machine Learning Applications in Cascading Failure Analysis in Power Systems: A Review
• Toward Dynamic Stability Assessment of Power Grid Topologies using Graph Neural Networks
• Power Failure Cascade Prediction using Graph Neural Networks
• Improving power-grid systems via topological changes, or how self-organized criticality can help stability
• Reinforcement Learning Optimizes Power Dispatch in Decentralized Power Grid
• Optimal location of reinforced inertia to stabilize power grids