Wind power prediction method and system based on small sample data

By segmenting the wind farm as a sub-region, terrain features are extracted, LSTM models are clustered and trained, and wind power prediction is carried out in combination with weather forecast data, the problem of wind power cluster power prediction under small sample data is solved, and efficient and accurate wind power prediction is achieved.

CN119691487AActive Publication Date: 2025-03-25STATE GRID SICHUAN ECONOMIC RES INST
4 Cites 0 Cited by

Patent Information

Application Number
CN202411833105.5
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-12-13
Publication Date
2025-03-25
Estimated Expiration
2044-12-13

AI Technical Summary

Technical Problem

The prior art is difficult to effectively predict regional wind power cluster power, especially under small sample data conditions, and the data scale is huge, and model training and prediction accuracy are difficult to meet the power scheduling needs.

Method used

By obtaining the topographic parameters, historical power generation data and historical wind direction data of the sample wind farm, the wind farm is divided into multiple sub-regions, the terrain characteristics of the sub-regions are extracted, the sub-region sample set is selected, the LSTM model is trained, and the wind power power prediction is carried out based on the weather forecast data.

Benefits of technology

Accurate prediction of wind power under small sample data conditions, reducing resource consumption for data storage and model training, and is suitable for various types of wind farms.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The invention discloses a wind power prediction method and system based on small sample data, and relates to the wind power and data processing technology, and the method comprises the steps: determining the power generation information of a fan corresponding to each sub-region according to the historical power generation data of a sample wind field, and dividing the sample wind field into a plurality of sub-regions according to the terrain parameters; topographic features of the sub-regions are extracted; clustering a terrain category; selecting a sub-region sample set based on each clustering center according to the relationship of the distance between each sub-region and the clustering center; for each clustering center, training an LSTM model based on the topographic features of the sub-region sample set and the power generation information of the corresponding wind turbine in the divided periods; and for a to-be-predicted target wind field, selecting the LSTM model trained by the corresponding clustering category to execute power generation prediction of the target wind field. Classification is constructed in combination with factors such as spatial positions and geographic conditions of the wind power plant, so that prediction of the wind power of the wind power plant under the small sample condition is realized.
Need to check novelty before this filing date? Find Prior Art