Two layer model predictive control for a cost efficient operation of island energy systems

Hybrid energy systems are a promising approach to achieving low cost energy supply in remote areas, comprising of diesel generators, storage devices and renewable sources. The main potential for cost reduction is through optimized operation of the island grid. Power management of these island systems can be challenging due to sudden variations in load demand and high fluctuations in power supplied by renewables. A control approach for efficient operation guaranteeing uninterrupted energy supply and satisfying all operation constraints, including detailed component modeling, is presented. The optimization approach is comprised of two control layers to calculate the optimal energy dispatch in the system and to further improve its robustness towards prediction errors. A model predictive control method, based on real time predictions of future power profiles, in combination with a boundary value problem for the online control of hybrid energy systems is proposed. The nonlinear mixed integer optimization problem is recursively solved using discrete dynamic programming. Finally, a simulation study is carried out to illustrate the performance and economic benefits of the proposed method.