Map technology-based electric power system state estimation model training method and apparatus

WO2025091683A1PCT designated stage expired Publication Date: 2025-05-08SOUTHERN POWER GRID DIGITAL GRID RESEARCH INSTITUTE CO LTD
View PDF 5 Cites 0 Cited by

Patent Information

Application Number
PCT/CN2023/143155
Authority / Receiving Office
WO · WO
Patent Type
Applications
Current Assignee / Owner
Priority Date
2023-10-30
Filing Date
2023-12-29
Publication Date
2025-05-08

AI Technical Summary

Technical Problem

The prior art is difficult to achieve accurate state estimation in multi-topological structure scenarios in power systems, and cannot adapt to topological changes, resulting in the safety and stability of power systems being affected.

Method used

The power system state estimation model based on graph technology is adopted. By obtaining the power data of candidate nodes and branches, the feature map generation layer, feature map weighting layer, learning layer and state parameter output layer are used for model training to generate the power system state estimation model. This model considers topological information between nodes and branches, improving the robustness and accuracy of state estimation.

Benefits of technology

It realizes accurate state estimation of multi-topological structure scenarios in the power system, can adapt to topological changes, and improves the safety and stability of the power system.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN2023143155_08052025_PF_FP_ABST
    Figure CN2023143155_08052025_PF_FP_ABST
Patent Text Reader

Abstract

A map technology-based electric power system state estimation model training method and apparatus. The method comprises: inputting node power data of candidate nodes in a target electric power system, and branch power data of candidate branches into a model to be trained, and generating a predicted voltage amplitude of the candidate nodes and a predicted phase angle difference of the candidate branches by means of sequential processing by a feature map generation layer, a feature map weighting layer, a first feature map learning layer, a second feature map learning layer, a third feature map learning layer, a feature map addition layer and a state parameter output layer of said model; and performing training on the basis of the predicted voltage amplitude and the predicted phase angle difference to obtain an electric power system state estimation model.
Need to check novelty before this filing date? Find Prior Art

Citation Information

Patent Citations

  • Power distribution network state evaluation method and system based on graph convolutional network

    CN113762625A

  • Rapid state estimation method based on graph neural network

    CN114221334A

  • Power system state estimation method based on message passing graph neural network

    CN115146538A

  • Power distribution network state rapid evaluation calculation method, system and device and storage medium

    CN116228465A

  • Spatio-temporal graph neural network for time series prediction

    US20230252285A1