Scholarlink

Predicting solar energy accurately is crucial for optimizing renewable energy systems. This study explores the application of machine learning techniques—namely Long Short-Term Memory (LSTM), Gated Recurrent Units (GRU), and Extreme Gradient Boosting (XGBoost)—to forecast solar energy output, focusing on Direct Normal Irradiance (DNI) based on meteorological features. Additionally, we investigate the effects of dimensionality reduction methods, including Independent Component Analysis (ICA) and Principal Component Analysis (PCA), on model performance. Our results reveal that both ICA and PCA improve prediction accuracy, with PCA slightly outperforming ICA in most cases. Among the models tested, LSTM and GRU exhibit superior performance in capturing both short- and long-term trends in solar irradiance, while XGBoost offers competitive results with simpler architectures.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Predicting Solar Energy Using LSTM, GRU, and XGBoost: An Analysis with ICA and PCA

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Hafid Ben Achour,
  • Ziani Said,
  • Youssef Zorgani,
  • Lokmane Ziani

摘要

Predicting solar energy accurately is crucial for optimizing renewable energy systems. This study explores the application of machine learning techniques—namely Long Short-Term Memory (LSTM), Gated Recurrent Units (GRU), and Extreme Gradient Boosting (XGBoost)—to forecast solar energy output, focusing on Direct Normal Irradiance (DNI) based on meteorological features. Additionally, we investigate the effects of dimensionality reduction methods, including Independent Component Analysis (ICA) and Principal Component Analysis (PCA), on model performance. Our results reveal that both ICA and PCA improve prediction accuracy, with PCA slightly outperforming ICA in most cases. Among the models tested, LSTM and GRU exhibit superior performance in capturing both short- and long-term trends in solar irradiance, while XGBoost offers competitive results with simpler architectures.