A method, system, device and storage medium for reconstructing photovoltaic power data

By combining the adaptive kernel function support vector regression model (AK-SVR) with the real-time physical conditions of the photovoltaic power plant, the problem of data loss or distortion caused by power curtailment and shutdown during photovoltaic power generation is solved, generating a high-precision, physically accurate photovoltaic power data sequence, which is suitable for automated data reconstruction of photovoltaic power plants.

CN122432504APending Publication Date: 2026-07-21CYG SUNRI CO LTD
View PDF 3 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-04-20
Publication Date
2026-07-21

AI Technical Summary

Technical Problem

Existing technologies fail to effectively integrate the nonlinear physical principles in the photovoltaic power generation process and cannot build adaptive models for data reconstruction, resulting in missing or distorted photovoltaic measured power data under conditions such as power curtailment and shutdown. Traditional methods cannot guarantee physical authenticity and accuracy.

Method used

A support vector regression model based on adaptive kernel function (AK-SVR) is adopted. By acquiring historical data of photovoltaic power plants, data preprocessing is performed using theoretical clear-sky irradiance curves and real-time physical conditions. The model is trained in different regions and combined with tilted surface irradiance and wind speed and temperature corrections to generate high-precision reconstructed power data.

Benefits of technology

It achieves continuous and complete power data reconstruction, and the reconstruction results are highly consistent with the actual meteorological conditions, improving the integrity and usability of the data. It is highly adaptable and accurate, and is suitable for automated data reconstruction of various photovoltaic power plants.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122432504A_ABST
    Figure CN122432504A_ABST
Patent Text Reader

Abstract

A photovoltaic power data reconstruction method, system, device and storage medium. The historical measured data of the photovoltaic power station is obtained, the effective power generation period is determined according to the theoretical clear sky horizontal irradiance curve, the effective data and the invalid data are marked, the training data set and the data set to be reconstructed are divided; for the two data sets, the standard slope total irradiance is calculated, and the real-time working temperature, wind speed and environmental temperature are combined for correction; the training data set after irradiance correction is partitioned, the support vector regression model based on adaptive kernel function is independently configured and trained for each irradiance area, and the corresponding sub-model of each irradiance area is output; the trained sub-model is called to predict the irradiance partition of the data in the data set to be reconstructed, the reconstructed power value sequence is calculated to replace the data in the data set to be reconstructed, and the reconstructed power data is obtained. The scheme can improve the authenticity and availability of the data, and provide more reliable support for subsequent analysis and decision-making.
Need to check novelty before this filing date? Find Prior Art

Citation Information

Patent Citations

  • Photovoltaic abnormal data reconstruction method and system based on feature weighted local abnormal factors

    CN119337069A

  • Distributed photovoltaic data reconstruction method and device, medium and equipment

    CN119557583A

  • Photovoltaic power station power data identification and reconstruction method and system

    CN120493110A