Scholarlink

<p>Accurate solar irradiance forecasting is critical for optimizing the design, control, and monitoring of solar energy systems, yet its intrinsic intermittency remains a major challenge. This study proposes and evaluates three novel hybrid deep learning models AME-LSTM, AME-CNN, and AME-GRU all integrating Attentive Meteo-Embedding (AME) with recurrent or convolutional architectures and trained using the Adam optimizer. Using hourly meteorological data (temperature, relative humidity, wind direction) from 2015 to 2023 in Pala, Chad, the models achieve high predictive accuracy across seasons. The AME-GRU model delivers the best overall performance, with a mean absolute error (MAE) of 10.45 W/m<sup>2</sup> in summer and an R<sup>2</sup> of 0.997. AME-CNN also performs strongly in summer and autumn (R<sup>2</sup> = 0.997), while AME-LSTM shows higher errors in winter (MAE = 24.88 W/m<sup>2</sup>, R<sup>2</sup> = 0.984). These results demonstrate that the proposed hybrid models significantly outperform conventional approaches and provide a robust foundation for solar energy forecasting. This work directly supports the development of a hybrid power plant in Pala, contributing to Chad’s energy transition.</p>

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

Application of hybrid deep learning models for seasonal solar irradiance forecasting in the city of Pala

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年7月25日

查看全文
  • Osée Mounkang,
  • Claude Vidal Aloyem Kaze,
  • Nzoko Tayo Dieudonné,
  • Hervice Roméo Fogno Fotso,
  • Paul Edouard Duclair Pountounynyi,
  • Germaine Kenmoe Djuidje,
  • Réné Tchinda

摘要

Accurate solar irradiance forecasting is critical for optimizing the design, control, and monitoring of solar energy systems, yet its intrinsic intermittency remains a major challenge. This study proposes and evaluates three novel hybrid deep learning models AME-LSTM, AME-CNN, and AME-GRU all integrating Attentive Meteo-Embedding (AME) with recurrent or convolutional architectures and trained using the Adam optimizer. Using hourly meteorological data (temperature, relative humidity, wind direction) from 2015 to 2023 in Pala, Chad, the models achieve high predictive accuracy across seasons. The AME-GRU model delivers the best overall performance, with a mean absolute error (MAE) of 10.45 W/m2 in summer and an R2 of 0.997. AME-CNN also performs strongly in summer and autumn (R2 = 0.997), while AME-LSTM shows higher errors in winter (MAE = 24.88 W/m2, R2 = 0.984). These results demonstrate that the proposed hybrid models significantly outperform conventional approaches and provide a robust foundation for solar energy forecasting. This work directly supports the development of a hybrid power plant in Pala, contributing to Chad’s energy transition.