الفهرس | Only 14 pages are availabe for public view |
Abstract The integration of renewable energy sources (RES) into the grid is a crucial step toward mitigating the energy crisis, advancing sustainable energy practices, and contributing to the decarbonization of the energy sector. This integration is imperative for meeting international climate goals and reducing the long-term environmental impact of power generation, thereby fostering a more sustainable and ecologically responsible future. However, this integration poses significant challenges to power systems’ inertia and frequency stability through the shift away from conventional synchronous generators to renewables-based systems. Microgrids are increasingly vital in integrating RESs into today’s power systems. They offer a localized, flexible solution for harnessing solar, wind, and other RESs, enhancing grid resilience and reducing carbon emissions. However, the shift towards renewables and the decentralized nature of microgrids introduce challenges, particularly low inertia. This reduction in inertia can affect the stability and reliability of the power supply, making it crucial to develop innovative solutions to mitigate these effects and ensure sustainable, stable energy distribution. This thesis addresses these challenges by proposing innovative solutions by designing and implementing virtual inertia controllers. The first part of this thesis introduces a virtual inertia controller using a high-pass filter (HPF) designed to maintain frequency and DC link voltage stability within microgrids during disturbances. The effectiveness of this controller is demonstrated through a comprehensive state-space linearized model, comparing its dynamic response with traditional low-pass filter (LPF)-based controllers and highlighting the significant impact of system parameters on stability. The microgrid structure under study comprises an AC microgrid with dynamic and static loads, a synchronous generator representing low inertia, and a DC microgrid with RESs, constant power loads, and resistive loads. The second part of the thesis introduces Artificial intelligence to the control methodology by integrating reinforcement learning controllers into the virtual inertia controllers. The thesis also presents a virtual synchronous generator (VSG) secondary controller employing the Deep Deterministic Policy Gradient (DDPG) reinforcement learning algorithm. This novel approach is meticulously tested against a conventional PI controller, showcasing its superior inertial response and robustness under varying load conditions. The thesis further explores the application of this controller in two distinct scenarios: a standalone power source within a microgrid and an integrated system complemented by a synchronous generator and RES. The VSG secondary controller provides essential inertia support and effectively restores frequency to its nominal values. In the final segment, the thesis examines a reinforcement learning (RL) based control algorithm that utilizes Twin Delayed Deep Deterministic Policy Gradient (TD3) and DDPG RL agents to induce virtual inertia in autonomous microgrids. The agents are trained on a linearized model and are assessed to enhance frequency stability, with results indicating a clear advantage over conventional LPF and HPF controllers. Subsequent testing in a nonlinear model environment confirms the resilience and adaptability of these agents across various operational scenarios. This comprehensive study confirms the effectiveness of RL in microgrid management. It contributes significantly to the discourse on stabilizing RES-integrated power systems, ultimately presenting a promising avenue for future research and development in energy sustainability. This thesis presents three pivotal studies that outline a multi-faceted approach to addressing the challenges posed by low inertia in RES-dominated microgrids. Through advanced control algorithms and reinforcement learning, the thesis demonstrates the viability of these innovative solutions in enhancing the stability and reliability of the evolving power grid. |