Nico Dabelstein, Managing Fluctuating Renewable Energy and Demand Using Microgrids

Nico Dabelstein was an IGCS research exchange grantee and a part of a research group led by IIT Madras, Chennai’s Prof. Dr. Krishna Vasudevan, where they’re pioneering the development of a cutting-edge microgrid on the campus area.

Nico’s focus on Deep Reinforcement Learning (DRL) aims to optimize energy distribution within the grid, employing this powerful machine learning technique to enhance effectiveness, environmental sustainability, and economic performance. Through meticulous data collection and Python modeling of microgrid components using libraries like stablebaselines3 and pandas, Nico’s team has crafted a sophisticated reward function considering costs, CO2 emissions, and battery state of charge. By integrating a Pandapower Python model for accurate component modeling and optimal power flow calculations, alongside an artificial neural network (ANN) for swift OPF approximations crucial for DRL agent training, they’re paving the way for a robust control algorithm to manage fluctuating renewable energy and demand within the microgrid.

You can read their full research brief below: