تعداد نشریات | 7 |
تعداد شمارهها | 399 |
تعداد مقالات | 5,376 |
تعداد مشاهده مقاله | 5,133,664 |
تعداد دریافت فایل اصل مقاله | 4,759,524 |
Designing a robust method to improve virtual inertia control performance in islanded microgrid | ||
AUT Journal of Electrical Engineering | ||
مقالات آماده انتشار، پذیرفته شده، انتشار آنلاین از تاریخ 16 دی 1403 اصل مقاله (1.12 M) | ||
نوع مقاله: Research Article | ||
شناسه دیجیتال (DOI): 10.22060/eej.2025.23655.5628 | ||
نویسنده | ||
Farhad Amiri* | ||
Department of Electrical Engineering, Tafresh University, Tafresh 39518-79611, Iran | ||
چکیده | ||
Power-electronic converters play a crucial role in the functioning of microgrids. However, these converters, characterized by their low inertia, present a significant challenge to maintaining a consistent frequency in islanded microgrids. To address this issue, an innovative concept known as virtual inertia control (VIC) has emerged as a promising solution for enhancing frequency stability in islanded microgrids. The VIC system does not perform well against disturbances and uncertainty related to microgrid parameters. Therefore, to overcome these problems, it needs a suitable controller in its structure. In this paper, a linear quadratic regulator (LQR) mode feedback controller based on deep learning is proposed to improve the performance of VIC in an islanded microgrid against disturbances and uncertainties in the system. The LQR controller uses measurements of system states and the integration of a deep network increases the accuracy and dynamic response of the feedback controller. This allows for fine-tuning of the control response, which exhibits significant robustness against uncertainty in system parameters and disturbances. To evaluate its effectiveness and compare it against alternative control approaches, comprehensive assessments have been conducted across multiple scenarios. The results indicate that the proposed method in the field of VIC surpasses previous approaches. | ||
کلیدواژهها | ||
Virtual inertia control؛ LQR؛ Deep learning؛ Performance | ||
آمار تعداد مشاهده مقاله: 34 تعداد دریافت فایل اصل مقاله: 66 |