Scholarlink

<p>The current study proposes a hybrid intelligent approach depending on support-vector-machine (SVM) for predicting the global tilted solar irradiation. The optimization of this model is carried out by combining the particle-swarm-optimization (PSO) and the grey-wolf-optimizer (GWO). This model is referred to be as HPSOGWO. The hybrid optimizer leverages the capability of the global search of PSO algorithm and the local exploitation strength of GWO for fine-tuning the parameters of SVM, enhancing prediction accuracy and robustness. Meteorological data is gathered from the Bouzareah region in Algeria, including the temperature, the relative humidity, the wind’s speed and its direction, the atmospheric pressure, and horizontal solar radiation, were used as input features. The performance of model and its effectiveness are assessed considering statistical metrices such as (1) the root-mean-squared error (RMSE), (2) mean-absolute-error (MAE), and (3) coefficient of correlation (R). The results demonstrated that the proposed hybrid particle-swarm with the grey-wolf optimized support-vector-machine (HPSOGWO–SVM) approach significantly outperforms conventional SVM optimization techniques, demonstrating its potential as an effective tool for accurate forecasting of the solar energy and the optimization of photovoltaic system design.</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

A hybrid particle swarm and grey wolf optimized support vector machine (SVM) for enhanced solar energy forecasting

  • 期刊论文

  • 发表日期：

    2026年8月3日

查看全文
  • Abdennasser Dahmani,
  • Ikram Kouidri,
  • Lotfi Mouni,
  • Abdel-Nasser Sharkawy

摘要

The current study proposes a hybrid intelligent approach depending on support-vector-machine (SVM) for predicting the global tilted solar irradiation. The optimization of this model is carried out by combining the particle-swarm-optimization (PSO) and the grey-wolf-optimizer (GWO). This model is referred to be as HPSOGWO. The hybrid optimizer leverages the capability of the global search of PSO algorithm and the local exploitation strength of GWO for fine-tuning the parameters of SVM, enhancing prediction accuracy and robustness. Meteorological data is gathered from the Bouzareah region in Algeria, including the temperature, the relative humidity, the wind’s speed and its direction, the atmospheric pressure, and horizontal solar radiation, were used as input features. The performance of model and its effectiveness are assessed considering statistical metrices such as (1) the root-mean-squared error (RMSE), (2) mean-absolute-error (MAE), and (3) coefficient of correlation (R). The results demonstrated that the proposed hybrid particle-swarm with the grey-wolf optimized support-vector-machine (HPSOGWO–SVM) approach significantly outperforms conventional SVM optimization techniques, demonstrating its potential as an effective tool for accurate forecasting of the solar energy and the optimization of photovoltaic system design.