부산대학교 도서관

Microgrids rely heavily on batteries for primary frequency regulation, but battery's state of charge (SOC) critically affects its ability to support frequency control. Efficient load frequency control (LFC) for isolated microgrids must considers battery's non-linear dynamics due to charge/discharge behavior and SOC variations, a factor often overlooked in existing research, which is essential for achieving reliable performance across all operating conditions. In this study, at first, stability analysis to assess how varying the initial SOC of the battery affects LFC operation in the isolated microgrids is conducted. Then a battery frequency support technique is proposed to manage the SOC within 20–80 % while regulating the frequency. Finally, an adaptive, coordinated LFC is developed to address challenges arise from the BESS's nonlinear characteristics. The proposed LFC is designed utilizing gain-scheduling technique and Real Coded Genetic Algorithm (RCGA) while its flexibility and effectiveness are validated through multiple operating scenarios adopting MATLAB simulation and hardware-in-the-loop. The results show that, during over- and under-frequency events and with different initial SOCs, the proposed controller enables the BESS to consistently and securely regulate the frequency within its allowed limits, while ensuring a smooth SOC recovery process as preparation for the subsequent events. For example, at under-frequency events, higher SOC levels significantly reduce the maximum ∆f, e.g. from 0.1745 Hz at 23.6 % SOC to 0.03035 Hz at 78 % SOC. During over-frequency events, while ∆f increases with higher SOC levels (e.g. from 0.03424 Hz at 30.86 % SOC to 0.1217 Hz at 69.29 % SOC), ∆f remains within its permissible range.