Optimizing Model Predictive Control for Frequency Stabilization in Marine Microgrids Using Hybrid Optimization Algorithms

Abstract Model predictive control (MPC) plays a vital role in maintaining frequency stability in marine microgrids, particularly as renewable energy sources (RESs) are increasingly integrated into maritime power systems. To address the challenges of variable generation and fluctuating loads, this study proposes a hybrid optimization framework that combines a genetic algorithm (GA) with Gorilla troop optimizer (GTO). The hybrid approach enhances MPC performance by improving reliability and efficiency in real-time frequency regulation. Developed in the matlab/simulink environment, the proposed GA-GTO-based MPC demonstrates improved computational efficiency and higher accuracy in frequency prediction. Simulation results indicate that the optimized controller reduces frequency oscillations from 1.2 Hz (proportional-integral-derivative (PID)) and 0.75 Hz (standard MPC) to 0.2 Hz, while also lowering response latency from 5 s to 2 s. These improvements highlight the potential of hybrid optimization techniques to advance control strategies for marine microgrids, ensuring stable operation in renewable energy–dominated environments. Future work will focus on adaptive real-time optimization using machine learning and scalability analysis for larger marine power systems.