A wind power grid-connected control method based on demand response

By collecting wind speed and meteorological data in remote island microgrids, and utilizing technologies such as long short-term memory networks and support vector machines, a multi-level distribution path and verification mechanism were constructed. This solved the data distribution and accuracy problems of wind power prediction, achieved efficient wind power grid connection control, improved prediction accuracy and scheduling efficiency, and ensured grid stability.

CN121461444BActive Publication Date: 2026-06-09HUANENG JIUQUAN WIND POWER CO LTD
View PDF 2 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
HUANENG JIUQUAN WIND POWER CO LTD
Filing Date
2025-10-22
Publication Date
2026-06-09

AI Technical Summary

Technical Problem

Existing wind power forecasting technologies are inadequate in terms of data distribution and forecast accuracy, making it difficult to meet the multi-level and multi-dimensional grid dispatching needs. This results in forecast data not being efficiently transmitted to appropriate control nodes, affecting the grid's dynamic response capability and grid connection stability.

Method used

By collecting real-time wind speed and meteorological data of microgrids on remote islands, a preliminary predicted output value is generated using a long short-term memory network. A multi-level distribution path is constructed by combining historical resource allocation records, the data accuracy is dynamically adjusted, a support vector machine is used to verify the deviation, and a gradient boosting decision tree is integrated to optimize scheduling efficiency, thus forming a continuously adaptable multi-level distribution mechanism.

Benefits of technology

It significantly improves the prediction accuracy and dispatch efficiency of microgrids under complex weather conditions, ensures grid connection stability, and is suitable for resource-constrained remote island scenarios.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121461444B_ABST
    Figure CN121461444B_ABST
Patent Text Reader

Abstract

The application provides a wind power grid connection control method based on demand response, comprising: obtaining a multi-level distribution path according to a preliminary predicted output value and a historical resource allocation record, and determining a hierarchical demand matching suitable for different time scales; if the hierarchical demand matching exceeds a preset threshold, optimizing the data accuracy with a low deviation threshold through a dynamic adjustment algorithm to obtain a refined prediction data set; extracting wind speed fluctuation characteristics from the refined prediction data set, verifying the system through a support vector machine model, and judging the prediction deviation degree; if, generating an output fluctuation correction coefficient according to the verification result to obtain a final grid connection stability index; backtracking the preliminary predicted output value according to the overall scheduling efficiency improvement scheme, obtaining updated meteorological data input, and obtaining a cyclically optimized predicted demand response; analyzing the hierarchical demand change from the cyclically optimized predicted demand response, judging whether further dynamic adjustment is needed, and obtaining a continuously adaptive multi-level distribution mechanism.
Need to check novelty before this filing date? Find Prior Art