Artificial Intelligence supported Power Quality Prediction and Mitigation - Adrian Eisenmann - Bog

This thesis introduces a fully data driven approach for the prediction and optimization of critical electrical grid states due to poor power quality. Therefore, a nonvolatile memory model for time series forecasting, designed to profit especially from big data bases and complex pattern use cases as well as an Artificial Intelligence based Smart Demand Side Management framework to enable system inherent resources / components for minimization of harmonic disturbances is applied to measured power grid scenarios.