An emergency resource demand prediction method and device based on WGAN-Stacking

By generating highly realistic fault samples using the WGAN-Stacking method and combining them with various quantitative indicators to filter data, and employing an ensemble learning framework, the problem of sample imbalance and complex coupling relationships in power grid emergency resource demand prediction was solved, achieving high-precision and stable resource allocation.

CN122366971APending Publication Date: 2026-07-10ELECTRIC POWER RES INST OF STATE GRID ZHEJIANG ELECTRIC POWER COMAPNY
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
ELECTRIC POWER RES INST OF STATE GRID ZHEJIANG ELECTRIC POWER COMAPNY
Filing Date
2026-04-10
Publication Date
2026-07-10

AI Technical Summary

Technical Problem

In existing technologies, the prediction of power grid emergency resource demand relies on historical case statistics and expert experience, which is highly subjective and difficult to cope with extreme icing disasters. Furthermore, single machine learning models have insufficient prediction accuracy and stability in the face of complex meteorological-geographical-power grid coupling relationships, especially under the problem of imbalanced samples, the model has insufficient prediction ability for minority classes.

Method used

A WGAN-Stacking-based approach is adopted to generate highly realistic fault samples through Wasserstein generative adversarial networks. The sample quality is screened by combining Fréchet Inception distance and kernel Inception distance. The Stacking ensemble learning framework is used to fuse heterogeneous base learners of RF, GBDT and SVM, and FNN is used for nonlinear combination to predict emergency resource demand.

Benefits of technology

It effectively expands the sample size of minority classes, improves the accuracy and stability of prediction models, and enables more accurate allocation of emergency resources to adapt to complex disaster scenarios.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122366971A_ABST
    Figure CN122366971A_ABST
Patent Text Reader

Abstract

This invention belongs to the field of resource allocation and discloses an emergency resource demand prediction method and apparatus based on WGAN-Stacking. The method includes: acquiring meteorological, geographical, and power grid data; generating fault samples using a Wasserstein generative adversarial network; quantifying the distribution difference between the generated fault samples and real samples based on FID distance and kernel Inception distance to screen target samples whose sample quality reaches a preset threshold, with the quantified distribution difference used to characterize the sample quality of the fault samples; inputting the target samples into a pre-constructed resource demand prediction model to predict the emergency resource demand in different regions; the resource demand prediction model adopts a Stacking ensemble learning framework, using RF, GBDT, and SVM as heterogeneous base learners and FNN as a meta-learner, learning the optimal nonlinear combination strategy between the prediction results of each base learner; and allocating emergency resources according to the prediction results output by the resource demand prediction model.
Need to check novelty before this filing date? Find Prior Art