Êtes-vous un étudiant de l'EPFL à la recherche d'un projet de semestre?
Travaillez avec nous sur des projets en science des données et en visualisation, et déployez votre projet sous forme d'application sur GraphSearch.
We consider low run-time complexity power management in distribution grids with renewable energy sources (RESs) and batteries, where forecasts are unavailable. We propose iterative Lyapunov Real-time Control (iLypRC), a fast online algorithm, which does not need forecasts, for the real-time battery control. iLypRC is designed via Lyapunov optimization and applies either to follow a previously computed dispatch plan or to optimize the monetary cost, while it satisfies look-ahead state-of-energy constraints. iLypRC requires only bounds on the uncertain quantities for each real-time interval. iLypRC efficiently accounts for grid losses, battery efficiency, and grid constraints via iterative linearizations of the power flow equations. We compute a theoretical upper bound on the difference between the cost of iLypRC and the cost of an oracle. Finally, numerical evaluations (using real data) on a 4-bus distribution grid and a 34-bus real Swiss grid show that iLypRC achieves a cost very close to that of a model predictive control (MPC) with a good forecast, but iLypRC needs no forecast and has much lower run-time complexity. Also, when the MPC forecast is inaccurate, iLypRC outperforms MPC.