Optimization Study of Photovoltaic Energy Output Prediction Model Using Transformer

DOI

10.2991/978-2-38476-553-9_45How to use a DOI?

Keywords

Transformer; photovoltaic power prediction; attention mechanism; deep learning; time series forecasting; smart grid

Abstract

Accurate photovoltaic (PV) power forecasting plays a crucial role in the reliable operation of smart grids and renewable energy scheduling. To address the challenges of nonlinear fluctuations and time-dependent variability in solar power generation, this study proposes an optimized Transformer-based prediction model for PV energy output. The model leverages a multi-head self-attention mechanism and positional encoding to capture both short-term and long-term dependencies between meteorological factors and historical power data.

Using real-world hourly datasets containing irradiance, temperature, wind speed, and humidity, the proposed Transformer model was compared with benchmark methods, including Backpropagation (BP) and Long Short-Term Memory (LSTM) networks. Experimental results demonstrate that the Transformer achieved the lowest Mean Absolute Error (MAE) and Root Mean Squared Error (RMSE), outperforming traditional neural networks in accuracy and robustness under dynamic weather conditions. Furthermore, attention-based feature analysis verified the interpretability of the model, aligning with the physical characteristics of PV systems. The proposed framework provides a practical and effective solution for intelligent solar power forecasting and smart grid energy management.

Copyright

© 2026 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - Ziqing He
PY  - 2026
DA  - 2026/03/25
TI  - Optimization Study of Photovoltaic Energy Output Prediction Model Using Transformer
BT  - Proceedings of the 2025 4th International Conference on Educational Science and Social Culture (ESSC 2025)
PB  - Atlantis Press
SP  - 394
EP  - 400
SN  - 2352-5398
UR  - https://doi.org/10.2991/978-2-38476-553-9_45
DO  - 10.2991/978-2-38476-553-9_45
ID  - He2026
ER  -