Abstract
Accuracy and simplicity issues of the widely-used aggregated model of wind farms focus on minimizing the error during the transient characteristics. In order to reduce the deviation between the detailed and reduced-order model, this paper presents a data-driven system identification method to discover a mathematical structure from the error data that possibly improve the established model. In this work, we devise a library of candidate dynamics tailored to wind farms for regression algorithm. This framework only requires fairly little data and is robust to noise, having good performance for the response to fast system variation. The simulation result illustrates the accuracy of the improved aggregation model of a wind power plant under low voltage ride-through (LVRT) mode.
Original language | English |
---|---|
Title of host publication | 2023 6th International Conference on Energy, Electrical and Power Engineering, CEEPE 2023 |
Number of pages | 6 |
Publisher | IEEE |
Publication date | 12 May 2023 |
Pages | 1219-1224 |
Article number | 10166207 |
ISBN (Print) | 979-8-3503-4828-6 |
ISBN (Electronic) | 979-8-3503-4827-9 |
DOIs | |
Publication status | Published - 12 May 2023 |
Event | 2023 6th International Conference on Energy, Electrical and Power Engineering (CEEPE) - China, Guangzhou Duration: 12 May 2023 → 14 May 2023 https://ieeexplore.ieee.org/abstract/document/10166207 |
Conference
Conference | 2023 6th International Conference on Energy, Electrical and Power Engineering (CEEPE) |
---|---|
Location | China |
City | Guangzhou |
Period | 12/05/2023 → 14/05/2023 |
Internet address |
Keywords
- Wind farm
- aggregation
- low voltage ride-through (LVRT)
- nonlinear dynamics
- symbolic regression