Scholarlink

<p>The renewable energy sector has grown substantially, driven by concerns over environmental pollution and rising fossil fuel prices. Solar energy holds the largest market share among renewable energy sources, making accurate solar energy prediction crucial in the face of changing weather patterns and climate impacts. Accurate solar energy prediction empowers decision-makers to plan better and optimize solar power generation, reducing reliance on fossil fuels and contributing to a cleaner, more sustainable energy future. This study analyzes various prediction models’ performance in predicting daily Solar Global Horizontal Irradiance (GHI) for Bhopal. The dataset spans from January 1, 2019, to January 1, 2023, providing a comprehensive evaluation of the models. This research examined five different models: Artificial Neural Network (ANN), Long Short-Term Memory (LSTM), Convolutional Neural Network (CNN), Prophet, and SARIMAX. Assessing the accuracy of these models using multiple error metrics, including Mean Absolute Error (MAE), Mean Squared Error (MSE), Root Mean Squared Error (RMSE), Mean Absolute Percentage Error (MAPE), and R<sup>2</sup> (Coefficient of Determination) allowed to make a comprehensive comparison. The findings revealed that the ANN model outperformed other models, achieving the best predictive performance for solar GHI in Bhopal. With an MAE of 0.1139&#xa0;kW-h/m<sup>2</sup>/day, MSE of 0.0191&#xa0;kW-h/m<sup>2</sup>/day, RMSE of 0.13848&#xa0;kW-h/m<sup>2</sup>/day, MAPE of 2.5466%, and R<sup>2</sup> of 0.9922, the ANN model demonstrated remarkable accuracy in predicting solar GHI.</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

Performance evaluation of statistical and deep learning models for daily solar global horizontal radiation prediction: implications for renewable energy and sustainability

  • 期刊论文

  • 发表日期：

    2025年3月7日

查看全文
  • Bharat Girdhani,
  • Meena Agrawal

摘要

The renewable energy sector has grown substantially, driven by concerns over environmental pollution and rising fossil fuel prices. Solar energy holds the largest market share among renewable energy sources, making accurate solar energy prediction crucial in the face of changing weather patterns and climate impacts. Accurate solar energy prediction empowers decision-makers to plan better and optimize solar power generation, reducing reliance on fossil fuels and contributing to a cleaner, more sustainable energy future. This study analyzes various prediction models’ performance in predicting daily Solar Global Horizontal Irradiance (GHI) for Bhopal. The dataset spans from January 1, 2019, to January 1, 2023, providing a comprehensive evaluation of the models. This research examined five different models: Artificial Neural Network (ANN), Long Short-Term Memory (LSTM), Convolutional Neural Network (CNN), Prophet, and SARIMAX. Assessing the accuracy of these models using multiple error metrics, including Mean Absolute Error (MAE), Mean Squared Error (MSE), Root Mean Squared Error (RMSE), Mean Absolute Percentage Error (MAPE), and R2 (Coefficient of Determination) allowed to make a comprehensive comparison. The findings revealed that the ANN model outperformed other models, achieving the best predictive performance for solar GHI in Bhopal. With an MAE of 0.1139 kW-h/m2/day, MSE of 0.0191 kW-h/m2/day, RMSE of 0.13848 kW-h/m2/day, MAPE of 2.5466%, and R2 of 0.9922, the ANN model demonstrated remarkable accuracy in predicting solar GHI.