الفهرس 
IntroductionOnly 14 pages are availabe for public view
 |  | from | 188 |  |  |
| | | from | 188 | | |
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.