Ultra long-Term Wind Farm Generation Forecast by Combining Numerical Weather Prediction with Gated Recurrent Units
| Autor: | Julia Penfield |
|---|---|
| Rok vydání: | 2021 |
| Předmět: |
Wind power
business.industry Computer science Astrophysics::High Energy Astrophysical Phenomena Numerical weather prediction Industrial engineering Term (time) Renewable energy Resource (project management) ComputerApplications_MISCELLANEOUS Physics::Space Physics Astrophysics::Solar and Stellar Astrophysics Operational planning Portfolio Dispatchable generation business Physics::Atmospheric and Oceanic Physics |
| Zdroj: | 2021 9th International Conference on Smart Grid (icSmartGrid). |
| DOI: | 10.1109/icsmartgrid52357.2021.9551245 |
| Popis: | Wind energy is an integral resource in the renewable energy portfolio of many utilities across the world. One of the challenges in planning for operations when dealing with wind farms is to account for the fact that wind power is not dispatchable, and forecasting of wind over a typical window of operational planning such as a 7-day long horizon is challenging. While most previous reports on forecasting wind power are focused on short term wind forecast, this study focuses on ultra-long term forecast of generation in wind farms by using physical models in combination with deep learning as a rapidly growing part of machine learning discipline. The proposed approach is compared against a traditional approach of wind power forecast using physical models and provides promising improvement in accuracy. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|