Robustness Analysis of AI Models in Critical Energy Systems

TL;DR

This paper evaluates how well AI models for power grid management maintain accuracy under the $N-1$ security criterion, revealing significant robustness issues when grid components disconnect, and underscores the importance of scenario-based training.

Contribution

It introduces a graph theory-based approach to analyze the robustness of AI models in critical energy systems under security constraints.

Findings

AI models lose accuracy after line disconnections.

Node connectivity significantly affects model robustness.

Practical scenario considerations are crucial for reliable AI deployment.

Abstract

This paper analyzes the robustness of state-of-the-art AI-based models for power grid operations under the $N-1$ security criterion. While these models perform well in regular grid settings, our results highlight a significant loss in accuracy following the disconnection of a line.%under this security criterion. Using graph theory-based analysis, we demonstrate the impact of node connectivity on this loss. Our findings emphasize the need for practical scenario considerations in developing AI methodologies for critical infrastructure.