Ultra-Short-Term Power Forecasting Method for Wind-Solar-Hydro Integration Based on Improved GRU-CNN
The models of wind, solar, and hydro energy systems are very different, and there are multiple uncertainties among them. High-precision power forecasting technology for wind, solar, and hydro is an important prerequisite for giving full play to the complementary characteristics of wind, solar, and h...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | zho |
| Published: |
State Grid Energy Research Institute
2023-09-01
|
| Series: | Zhongguo dianli |
| Subjects: | |
| Online Access: | https://www.electricpower.com.cn/CN/10.11930/j.issn.1004-9649.202209120 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | The models of wind, solar, and hydro energy systems are very different, and there are multiple uncertainties among them. High-precision power forecasting technology for wind, solar, and hydro is an important prerequisite for giving full play to the complementary characteristics of wind, solar, and hydro. To this end, an integrated ultra-short-term power forecasting method is proposed based on gated recurrent units (GRUs) and convolutional neural networks (CNNs), which can consider the temporal and spatial correlation characteristics of heterogeneous energy sources. Firstly, the correlation characteristics of different data of different stations in the area are analyzed, and then, by introducing a temporal attention mechanism, the mapping relationship between historical meteorological/power data and future power data is established based on the improved GRU-CNN network, which realizes the multi-station integrated ultra-short-term forecasting. The calculation example results show that the forecasting method proposed in this paper can realize the integrated high-precision ultra-short-term power forecasting of regional wind, solar, and hydro power stations, and the model effect is better than the single-field forecasting method and general integrated forecasting method, with higher modeling efficiency. |
|---|---|
| ISSN: | 1004-9649 |