Masterarbeit: Anwendung physikinformierten maschinellen Lernens für die Regelung netzgekoppelter Wechselrichter (Betreuung: Prof. Theis) (w/m/div.)

Develop a physics-informed current controller for grid-connected inverters by evaluating linear and predictive regulators as benchmarks. Research current AI-based implementations in literature and design a concept to improve AI-driven black-box approaches while reducing their application drawbacks. Select an AI concept incorporating machine learning, recurrent neural networks, or reinforcement learning, and model the system with the implemented regulator if needed. This six-month Master's thesis requires prior university enrollment and involves office and lab work with occasional remote options.