Development of a Python model with a scheduling algorithm for smart grid energy management

The shift toward renewable and sustainable energy systems and the growing complexity of modern electricity networks have propelled the development of smart grids: electricity networks that use digital technologies to better match the supply and demand of electricity in real-time, while minimizing costs. This thesis inquires into an optimization-based approach for managing a hybrid main grid and Peer-To-Peer (P2P) electricity market in a network of prosumers, i.e., energy consumers who can also produce electricity. The focus is on solving the Economic Dispatch (ED) problem through convex optimization, enabling efficient power generation, storage and trading operations. A scenario of a small network of three prosumers is simulated in Python, followed by a scaled-up model of thirty prosumers. A fixed scheduling algorithm is designed in order to simulate the power management operations. The model is then optimized using the CVXPY library. The results show significant reductions in operational costs, highlighting the effectiveness and the efficiency of the optimization strategy. Further, the model is successfully deployed in a ROS 2 environment, replicating the Python simulation while enabling real-time communication. This study provides valuable insights into the scalability and efficiency of optimization strategies in decentralized energy systems.