Scholarlink

<p>Solar panels play a crucial role in the transition toward sustainable energy and have significant environmental, economic, and social implications. Maximum temperature and maximum wind speed significantly influence the efficiency and performance of solar panels. Understanding these factors is essential for optimizing solar energy systems. This study develops a hybrid deep learning model to accurately forecast key meteorological parameters. By analyzing and predicting two important parameters, maximum temperature and maximum wind speed, monthly, this system gives warnings and necessary functions to designers and implementers to improve and increase efficiency. In order to model the 80:20 split, a model was created for the years 2000 to 2023. Given the lack of other meteorological parameters, four lags were used for max temperature and max wind speed in this study to avoid both complexity and the scarcity of sufficient other parameters. By merging Convolutional Neural Network (CNN), Gated Recurrent Unit (GRU), and Long Short-Term Memory (LSTM) models, this study demonstrates an effective system for examining 12 significant solar parks across the globe. The results showed that, on average, the CNN-GRU-LSTM hybrid model had a root mean square error (RMSE) of 1.41&#xa0;°C for maximum temperature and 1.97&#xa0;km/h for maximum wind speed, with coefficients of determination (R²) of 0.97 and 0.96, respectively. The model outperforms other models in reducing errors by 57% and 44% for temperature and wind speed, respectively. Stability was verified over repeated runs with standard deviations of 0.10–0.16&#xa0;°C; 0.15–0.28&#xa0;km/h. By applying the Faiman model, these predictions are converted into measurable power losses (e.g., 11% losses in warm, low-wind conditions), allowing for proactive cooling and wind storage measures. Given the increasing progress in the field of renewable energy, including solar panels, this approach, with high accuracy in predicting two important parameters, can improve the performance of solar panels in response to changing weather conditions.</p> Graphical Abstract <p></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

Meteorological prediction for improving the performance of solar panels by a hybrid CNN-GRU-LSTM model: large-scale solar parks

  • 期刊论文

  • 发表日期：

    2026年7月13日

查看全文
  • Erfan Abdi,
  • Mohammad Ali Ghorbani,
  • Debasmita Misra,
  • Adeyemi Olusola,
  • Kehinde Adeyeye,
  • Saeid Kargar

摘要

Solar panels play a crucial role in the transition toward sustainable energy and have significant environmental, economic, and social implications. Maximum temperature and maximum wind speed significantly influence the efficiency and performance of solar panels. Understanding these factors is essential for optimizing solar energy systems. This study develops a hybrid deep learning model to accurately forecast key meteorological parameters. By analyzing and predicting two important parameters, maximum temperature and maximum wind speed, monthly, this system gives warnings and necessary functions to designers and implementers to improve and increase efficiency. In order to model the 80:20 split, a model was created for the years 2000 to 2023. Given the lack of other meteorological parameters, four lags were used for max temperature and max wind speed in this study to avoid both complexity and the scarcity of sufficient other parameters. By merging Convolutional Neural Network (CNN), Gated Recurrent Unit (GRU), and Long Short-Term Memory (LSTM) models, this study demonstrates an effective system for examining 12 significant solar parks across the globe. The results showed that, on average, the CNN-GRU-LSTM hybrid model had a root mean square error (RMSE) of 1.41 °C for maximum temperature and 1.97 km/h for maximum wind speed, with coefficients of determination (R²) of 0.97 and 0.96, respectively. The model outperforms other models in reducing errors by 57% and 44% for temperature and wind speed, respectively. Stability was verified over repeated runs with standard deviations of 0.10–0.16 °C; 0.15–0.28 km/h. By applying the Faiman model, these predictions are converted into measurable power losses (e.g., 11% losses in warm, low-wind conditions), allowing for proactive cooling and wind storage measures. Given the increasing progress in the field of renewable energy, including solar panels, this approach, with high accuracy in predicting two important parameters, can improve the performance of solar panels in response to changing weather conditions.

Graphical Abstract