Day-Ahead Photovoltaic Power Forecasting Using Multi-Source Data Fusion

Maoyuan Li; Sudi Xu; Mijia Song

doi:10.54097/1554g234

Authors

Maoyuan Li
Sudi Xu
Mijia Song

DOI:

https://doi.org/10.54097/1554g234

Keywords:

Photovoltaic Power, Day-Ahead Forecasting, Numerical Weather Prediction, Spatial Downscaling, Deep Learning.

Abstract

This study proposes a day-ahead forecasting framework for photovoltaic power generation, integrating numerical weather prediction and deep learning to enhance prediction accuracy and optimize grid scheduling. Using 15-minute resolution data from August 2018 to June 2019, a comprehensive model was developed to analyze PV power characteristics, extracting seasonal and diurnal patterns via solar irradiance theory and Fast Fourier Transform. An Attention-based Long Short-Term Memory model, combined with NWP data, predicts 24-hour power curves with a mean absolute error of 5.2 kW. Further enhancements using Convolutional Neural Networks and Bidirectional LSTM, alongside spatial downscaling with XGBoost and Kriging interpolation, improve accuracy in complex terrains, reducing wind speed prediction MAE by 69.93%. The framework performs robustly across sunny, cloudy, and extreme weather scenarios, offering reliable support for grid operations and renewable energy integration.

References

[1]Shi Weiguo, Chen Yanwen, Zhou Liang. Simulation study of wind power generation power prediction based on deep learning[J/OL]. Power Electronics Technology, 1-8[2025-05-26].

[2]HUANG Li, GAN Hengyu, LIU Xingju, etal. Ultra-short-term power prediction of photovoltaic power generation system based on deep learning[J]. China New Technology and New Products, 2025, (09): 4-6.

[3]H.Y. Yan. Research on power prediction of photovoltaic power generation by integrating PSO and BP neural network[J]. Electrical Technology and Economy, 2025, (04): 22-24.

[4]XIE Weichong, WANG Yonghong, ZHANG Kaixuan. Analysis of ultra-short-term power prediction methods and application scenarios for photovoltaic power generation[J]. Electric Power Safety Technology, 2025, 27(04): 38-41.

[5]LI Bo, GUO Xiaoqiang, QIAO Shangyou, etal. Simulation of power generation prediction under multi-variable load conditions in new energy power plants[J]. Computer Simulation, 2025, 42(04): 69-72+98.

[6]CHANG Zeyu, TIAN Liang. Ultra-short-term prediction of photovoltaic power generation based on Bi-LSTM and Kalman[J]. China Test, 2025, 51(05): 141-147.

[7]LI Chenghao, YANG Yongbiao, SONG Jiaqi, etal. Short-term power prediction of photovoltaic power generation in terrace area based on IMVMD and BiLSTM-SARIMA combined model[J]. Journal of Solar Energy, 2025, 46(02): 433-440.

[8]Lan Z, Wu JJ. Short-term photovoltaic power prediction based on improved particle swarm optimization BP neural network[J]. Technology and Innovation, 2025. 04. 002

[9]Yang Xiaoyu, Duan Ping, Yang Xiao, etal. Application of dynamic echo state network in short-term power prediction of photovoltaic power generation[J]. Electroacoustic Technology,2025,49(02):61-63.

[10]Wang, Xiaomei. Research on the application of artificial intelligence methods in photovoltaic power prediction[J]. China Informatization, 2025, (02): 72-74.