DeepAI AI Chat
Log In Sign Up

Contingency Analyses with Warm Starter using Probabilistic Graphical Model

by   Shimiao Li, et al.

Cyberthreats are an increasingly common risk to the power grid and can thwart secure grid operation. We propose to extend contingency analysis (CA) that is currently used to secure the grid against natural threats to protect against cyberthreats. However, unlike traditional N-1 or N-2 contingencies, cyberthreats (e.g., MadIoT) require CA to solve harder N-k (with k >> 2) contingencies in a practical amount of time. Purely physics-based solvers, while robust, are slow and may not solve N-k contingencies in a timely manner, whereas the emerging data-driven alternatives to power grid analytics are fast but not sufficiently generalizable, interpretable, or scalable. To address these challenges, we propose a novel conditional Gaussian Random Field-based data-driven method that is bothfast and robust. It achieves speedup by improving starting points for the physical solvers. To improve the physical interpretability and generalizability, the proposed method incorporates domain knowledge by considering the graphical nature of the grid topology. To improve scalability, the method applies physics-informed regularization that reduces the model size and complexity. Experiments validate that simulating MadIoT-induced N-k contingencies with our warm starter requires 5x fewer iterations for a realistic 2000-bus system.


page 1

page 8


GridWarm: Towards Practical Physics-Informed ML Design and Evaluation for Power Grid

When applied to a real-world safety critical system like the power grid,...

Probabilistic Graphs for Sensor Data-driven Modelling of Power Systems at Scale

The growing complexity of the power grid, driven by increasing share of ...

Bounding Data-driven Model Errors in Power Grid Analysis

Data-driven models analyze power grids under incomplete physical informa...

Knowledge- and Data-driven Services for Energy Systems using Graph Neural Networks

The transition away from carbon-based energy sources poses several chall...

Physics-Informed Deep Neural Network Method for Limited Observability State Estimation

The precise knowledge regarding the state of the power grid is important...

Power Grid Behavioral Patterns and Risks of Generalization in Applied Machine Learning

Recent years have seen a rich literature of data-driven approaches desig...

Multi-Source Data-Driven Outage Location in Distribution Systems Using Probabilistic Graph Learning

Efficient outage location is critical to enhancing the resilience of pow...