Genetic algorithm-based control strategies for efficient residential DC microgrid systems with hybrid storage

Abstract

This paper presents a power management and control strategy for a residential DC microgrid (DCMG) incorporating photovoltaic (PV) systems, fuel cells (FCs), and a hybrid energy storage system (HESS). The fluctuations in the DC bus voltage, arising from intermittent PV generation and variable load conditions, are mitigated by the HESS, which comprises both batteries and supercapacitors (SCs). This control strategy adopts, batteries to handle slow-frequency power surges, whereas SCs are employed to manage rapid frequency
fluctuations effectively. The proposed controllers are optimized using an evolution-based Genetic Algorithm (GA), eliminating the need for extensive mathematical modeling of the system. Comparative analysis between the GA-tuned and conventionally tuned controllers is conducted based on performance metrics, including overshoot, undershoot, and settling time. The simulation results indicate that the proposed controller performs satisfactorily, achieving a maximum overshoot of 3.08%, a maximum undershoot of 2.95%, and a settling time of 44.5 ms. To further assess the efficacy and robustness of the controllers, they are subjected to disturbances in sensor readings and variations in system parameters within a range of ±25 % of their nominal values. Additionally, to validate the practical applicability of the proposed system, the simulation results are corroborated using a real-time FPGA-based simulator (OP 5700).