Power virtual energy storage control method based on multi-energy complementation

By adaptively adjusting the learning rate of the XGBoost model, the problem of insufficient prediction in power virtual energy storage systems caused by traditional fixed learning rates is solved, achieving more efficient energy dispatch and response speed.

CN120855429AActive Publication Date: 2025-10-28GANZHOU TAO ROOF PHOTOVOLTAIC POWER GENERATION EQUIPMENT CO LTD
View PDF 7 Cites 0 Cited by

Patent Information

Application Number
CN202510956233.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-07-11
Publication Date
2025-10-28
Estimated Expiration
2045-07-11

Smart Images

  • Figure CN120855429A_ABST
    Figure CN120855429A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of electric power information, in particular to an electric power virtual energy storage control method based on multi-energy complementation, and the method comprises the steps: obtaining an adjustment direction factor of a learning rate after the end of any round of iteration according to a loss function in an XGBoost model in a process of predicting an electric power load through employing the XGBoost model; according to the historical data distribution condition of each key feature in any round of iteration and the prediction effect after any round of iteration is finished, the comprehensive adjustment amplitude of the learning rate after any round of iteration is finished is obtained; and according to the adjustment direction factor and the comprehensive adjustment amplitude, obtaining an adaptive learning rate after any round of iteration is finished, and according to the adaptive learning rate after each round of iteration is finished, obtaining a final prediction value of the power load output by the XGBoost model, thereby improving the accuracy of predicting the power load by using the XGBoost model, and improving the prediction efficiency of the power load. And the operation efficiency of the electric power virtual energy storage system is improved.
Need to check novelty before this filing date? Find Prior Art

Citation Information

Patent Citations

  • Model parameter training method, device and system

    CN104346629A

  • Electric power system short-term load prediction method based on GEP parameter optimization XGBoost

    CN111340273A

  • Model training method and device, equipment and medium

    CN119227836A

  • A user load forecasting system based on adaptive learning of historical load data

    CN119783865A

  • Building air conditioner cooling load prediction method based on RFE-VaDE-XGBoost feature selection and HN-Adam-STGAT

    CN120197754A