Hybrid Dollmaker optimisation algorithm and a simple and efficient metapath aggregated network for EV power management and voltage regulation in DC microgrids

Efficient power management and voltage regulation are critical for the stability of Electric Vehicle (EV)-integrated DC Microgrids (MGs), especially when incorporating Renewable Energy Sources (RESs) such as wind and battery storage. These systems must optimise energy usage while managing the charging and discharging of EVs. However, the intermittency of RESs and fluctuating load demands often lead to power outages and voltage instability. To address these challenges, this work proposes a novel hybrid method, DmOA-SEMAN, combining the Dollmaker Optimisation Algorithm (DmOA) and the Simple and Efficient Metapath Aggregated Network (SEMAN). The DmOA component optimally schedules EV charging and discharging, ensuring improved power balance and enhanced voltage stability. SEMAN facilitates proactive control by accurately forecasting dynamic load requirements and voltage variations. The hierarchical framework is implemented in MATLAB and benchmarked against advanced models, including GAN-MAPPO, MARL-MIO, PCGAN-GJO, ABC-DLNN, FLC-ANN, ANN-AOA, and MLP. Simulation results demonstrate that DmOA-SEMAN achieves a system efficiency of 98.8% while reducing power losses to 32.67 W. This intelligent hybrid strategy effectively stabilises voltage, minimises inefficiencies, and supports reliable energy management (EM) in EV-integrated DC MGs, promoting seamless integration with renewable energy systems. The approach offers a practical and safe solution for enhancing grid stability and EV energy coordination.