Machine Learning for Power Grid Control: A Project for Enhancing Resilience through Data Quality

Abstract

Machine learning methods offer promising capabilities for forecasting, decision support, and automated control, but their trustworthiness in critical infrastructures depends heavily on the quality of the underlying data. This project addresses these challenges by systematically analyzing and integrating data quality considerations into the entire ML lifecycle. The project’s outcomes include a dedicated data quality framework that identifies leverage points where interventions can enhance the reliability of ML models. By explicitly embedding data quality into the design and application of AI methods, this project contributes to strengthening trustworthiness and resilience in critical energy infrastructures.