A novel energy management system based on two-level hierarchical economic model predictive control for use in microgrid control

• Design of an efficient Two-Level Hierarchical Economic MPC for microgrid energy management. • Detailed modeling of energy storage components in microgrids. • Step-by-step design of the overall microgrid control architecture. • Simulations validate the superior performance of the proposed economic bi-level controller. This study proposes an innovative energy management system (EMS) based on two-level hierarchical model predictive control, designed for microgrid control. MPC-based EMS are economically efficient and beneficial, although the real-time implementation often requires hierarchical structures due to high computational costs. The high level defines a long-term economic optimisation, with a longer sampling period and prediction horizon. Traditionally, the lower level is based on a tracking index of the high-level reference with a shorter sampling period and prediction horizon. In this case, it is necessary to determine the optimisation index weights, a complex selection that depends on the specific characteristics of each microgrid. Generally, the weights are determined by trial and error, lacking a clear physical meaning. Therefore, in this work, a new approach is proposed where the low level performs economic optimisation (similar to high level) in a short-term. Both approaches were evaluated through a practical case in a microgrid. The results obtained show that the traditional tracking approach can generate undesirable effects (unnecessary energy transactions and frequent switching of devices) that produces economic losses, which can only be mitigated by a careful adjustment of the weights. In contrast, the proposed economic approach eliminates these undesirable behaviours, and the need for weight definition, achieving an economic improvement of 3–25%. Therefore, the results confirmed the effectiveness of the proposed new approach, which provides better economic results while overcoming the limitations of traditional tracking methods.