Short-term photovoltaic power prediction method based on weather classification

By combining the fuzzy C-mean clustering algorithm, long-term and short-term memory network and empirical wavelet transformation, photovoltaic power prediction is carried out for different weather types, and the problems of poor anti-interference ability and low prediction accuracy in the existing technology are solved, and the problem of poor anti-interference ability and low prediction accuracy is achieved. Higher prediction accuracy and robustness.

CN119009957APending Publication Date: 2024-11-22STATE GRID JIANGSU ELECTRIC POWER CO XUZHOU POWER SUPPLY CO
View PDF 0 Cites 2 Cited by

Patent Information

Application Number
CN202411034659.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-31
Publication Date
2024-11-22

AI Technical Summary

Technical Problem

The existing short-term prediction methods for photovoltaic power generation have poor anti-interference ability under extreme weather conditions, and relying solely on historical data analysis leads to low prediction accuracy, and lack prediction methods based on weather classification.

Method used

Using a combined prediction model of fuzzy C-mean clustering algorithm (FCM), long and short-term memory network (LSTM) and empirical wavelet transform (EWT), we can screen weather variables through Pearson's correlation coefficient to perform weather classification and independent prediction, and improve photovoltaics Accuracy and robustness of power prediction.

Benefits of technology

It effectively avoids the impact of extreme weather on the overall prediction model, improves the accuracy and robustness of short-term photovoltaic power prediction, reduces prediction errors, and enhances the prediction ability of different seasons and weather types.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119009957A_ABST
    Figure CN119009957A_ABST
Patent Text Reader

Abstract

The invention discloses a short-term photovoltaic power prediction method based on weather classification, and the method comprises the following steps: obtaining meteorological variables affecting photovoltaic output power through calculating a Pearson's correlation coefficient, and screening out the meteorological variables with high correlation as key features; combining the screened key features into vectors, performing clustering analysis on the vectors by adopting an FCM method, and classifying results according to weather types; for different weather types, an EWT-LSTM combined prediction model is adopted to train the prediction module unit; and determining the weather type of the prediction day according to the meteorological data before the prediction day, taking the historical photovoltaic data and the meteorological data before the prediction day as input, and importing the historical photovoltaic data and the meteorological data before the prediction day into the EWT-LSTM combined prediction model to obtain a photovoltaic power prediction result of the prediction day. According to the invention, short-term photovoltaic power generation power prediction is carried out based on the combined prediction model combining the FCM, the LSTM and the EWT, and the precision of the prediction result is high.
Need to check novelty before this filing date? Find Prior Art

Citation Information

Cited By

  • Distribution area distributed photovoltaic power prediction method and system

    CN120109786A

  • Photovoltaic power station day-ahead generation power prediction method fusing numerical weather forecast information and scene perception strategy

    CN121808690A