Learning-Based Virtual Inertia Control of an Islanded Microgrid With High Participation of Renewable Energy Resources

Abstract

Renewable energy sources (RESs) are increasingly used to meet consumer demands in microgrids (MGs). However, high RES integration introduces system frequency stability, inertia, and damping reduction challenges. Virtual inertia (VI) control has been recognized as an effective solution to improve system frequency response in such circumstances. Conventional control techniques for VI control, which rely heavily on specific operating conditions, can lead to flawed performance during contingencies due to their lack of adaptivity. To address these challenges, this article proposes a novel attitude found on brain emotional learning (BEL) to emulate VI and damping for effective frequency control. The BEL-based controller is capable of quickly learning and handling the complexity, nonlinearity, and uncertainty intrinsic to the MGs, and it operates independently of prior knowledge of the system model and parameters. This characteristic enables the controller to adapt to various operating conditions, improving its robustness. The simulation results across three disturbance scenarios show that the proposed BEL-based controller significantly improves the system's response. The absolute maximum deviation of frequency was reduced to 0.0561 Hz in the final scenario, marking performance enhancements of 46.62% and 49.04% when compared with the artificial neural network-based proportional–integral control and the standard proportional control, respectively. This underlines the controller's adaptability and superior effectiveness in varying operating conditions.