A novel energy management system based on move-blocking based predictive control for use in microgrid control

Abstract

[EN] The increasing complexity of renewable-based microgrids demands advanced Energy Management Systems (EMS) capable of efficiently managing multi-time scale dynamics. Conventional Model Predictive Control (MPC) based EMS often rely on complex hierarchical structures with problematic weighting factors, which can cause undesirable effects, leading to suboptimal performance. This work proposes and validates an innovative single-layer EMS that leverages a move-blocking strategy. The key novelty lies in embedding long-term economic objectives and short-term physical constraints into a unified optimisation problem, eliminating the need for hierarchical layers and manual weight tuning. The performance of the proposed EMS was benchmarked against a conventional bilevel hierarchical EMS and an optimal reference case. The results demonstrate significant quantitative and qualitative improvements. Our approach reduces total operating costs by over 8%. This saving is composed of several key factors: a 13 % decrease in degradation-related costs attributable to more stable management of the hydrogen system, a reduction in fixed operating costs by minimising the usage time of the hydrogen systems, and a 7 % reduction in variable energy costs. Computationally, the strategy is highly efficient, reducing computation time by 99 % compared to the optimal case and confirming its real-time feasibility. Furthermore, it ensures more stable operation of the hydrogen system and increases system autonomy. These findings position the proposed move-blocking MPC framework as a scalable, robust, and computationally efficient solution for intelligent energy management in next-generation microgrids.