Optimal power dispatching strategies in smart-microgrids with storage

With the development of decentralized power sources based on renewable energy, power grids need smarter operations to be run properly. This paper investigates different procedures for the optimal power dispatching of a grid-connected prosumer with an energy storage consisting in a high speed flywheel. An off-line optimal scheduling for the day ahead aims at minimizing the cost with regards to the daily energy rates and considering the forecasts for both consumption and production. That dispatching is performed thanks to global optimization procedures based on a trust-region method or on a niching genetic algorithm. Another approach using step by step optimization and exploiting an original self-adaptive dynamic programming strategy is also developed. The paper discusses the performance of all the considered methods with regards to the obtained results and the computational time.