Lightweight multi-source data fusion photovoltaic power generation power prediction method and system

Through the combination of multi-source data fusion and BiGRU-Attention network, the limitations of the existing photovoltaic power generation prediction technology and the heavy computing burden are solved, and efficient and accurate photovoltaic power prediction is achieved.

CN119691685AActive Publication Date: 2025-03-25GUANGDONG UNIV OF TECH
5 Cites 2 Cited by

Patent Information

Application Number
CN202510176916.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-02-18
Publication Date
2025-03-25
Estimated Expiration
2045-02-18

AI Technical Summary

Technical Problem

The existing photovoltaic power prediction technology has the problems of limitations of a single data source, heavy computing burden, high hardware resource requirements, and poor adaptability to complex weather conditions and nonlinear characteristics of cloud cluster movement.

Method used

The lightweight multi-source data fusion method is adopted to obtain satellite cloud maps, foundation cloud maps, meteorological data and photovoltaic power historical data, and dynamic cloud tracking method, multi-scale frequency domain optimization cloud segmentation method and mutual information method to extract features, and build a BiGRU-Attention network for multi-source data fusion and prediction.

Benefits of technology

It improves data utilization efficiency and prediction accuracy, reduces computing burden and model resource requirements, is suitable for deployment in low-resource scenarios, and realizes accurate and lightweight power prediction.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The invention relates to the technical field of photovoltaic power generation power prediction, in particular to a lightweight multi-source data fusion photovoltaic power generation power prediction method and system. According to the method, the satellite cloud picture, the foundation cloud picture, the key meteorological factors and the historical photovoltaic power generation power data are combined, the data utilization efficiency and the prediction precision can be improved, an optimized feature extraction method is applied, the calculation speed is increased, the calculation burden is reduced, a lightweight architecture combining traditional time sequence modeling and an attention mechanism is used, and the prediction efficiency is improved. The method is suitable for deployment of low-resource scenes; according to the method, feature extraction and fusion of photovoltaic power generation multi-source data are realized, and accurate and lightweight power generation power prediction is realized, so that a manager of a small centralized photovoltaic power station in a rural area can perform real-time monitoring and accurate and efficient prediction on the photovoltaic power generation power on a mobile terminal; therefore, a corresponding energy management decision is made.
Need to check novelty before this filing date? Find Prior Art