Smart energy coordination in microgrid clusters using hybrid model predictive control and differential evolution optimization

The increasing penetration of renewable energy technologies (RETs) and energy storage systems (ESSs) into modern power grids requires advanced energy management strategies to ensure reliability and sustainability. However, a critical research gap remains in simultaneously achieving global economic optimality and fast dynamic regulation. Conventional static approaches often struggle to balance the non-linear complexity of multi-objective economic dispatch against the strict real-time stability constraints imposed by intermittent generation. This paper proposes a novel energy management system (EMS) that combines model predictive control (MPC) with a differential evolution (DE) optimization algorithm (MPC-DE EMS) to enable the real-time, multi-objective control of MG clusters (MGC). Implemented within the IEEE 15-bus distribution network, the control architecture employs a reduced-order MPC with only two inputs and outputs, achieving four concurrent objectives: (i) total independence from the main utility grid, (ii) minimization of operating costs, (iii) reduction of power generation losses, and (iv) maximization of RETs utilization. The DE algorithm computes a multi-criteria optimal cost reference incorporating operational costs, generation losses, and penalties for underutilized renewable capacity, which guides the MPC to achieve an optimal economic power dispatch. Simulation results demonstrate the superiority of the proposed framework: the MPC-DE EMS reduces the total operating cost by 34.78% (from 29.59 €/h to 19.3 €/h) compared to a conventional proportional control. Furthermore, when benchmarked against a particle swarm optimization (PSO) based strategy, the proposed method achieved a 32.19% improvement in economic costs and an 11.63% reduction in power losses. Additionally, RETs utilization increased by 9.89%, validating the framework’s capacity to simultaneously improve economic efficiency and operational performance. The control strategy was further validated in real time using a hardware-in-the-loop (HIL) platform based on an OPAL-RT4512 and a Siemens SIMATIC S7-1500 PLC, confirming its robustness and practical feasibility for advanced MG coordination.