Scholarlink

<p>This study comprehensively analyses photovoltaic (PV) system performance by examining its characteristic curves and conducting a comparative evaluation using three methodologies: experimental investigation, artificial neural network (ANN) modelling, and physical simulation, which constitutes the study’s novelty. Recognizing research gaps such as data limitations, incomplete models, ANN applicability, measurement inaccuracies, and the need for validated experimental inputs in PV energy harvesting, an outdoor experimental setup, integrated with LabVIEW software, was employed to collect voltage and current data from a PV module under real-world conditions. Based on these data, an ANN model was developed to predict power generation, achieving a high regression value of 0.9963, indicating a strong correlation with measured results. Furthermore, the ANN model was utilized for maximum power point tracking (MPPT) to optimise power output under varying environmental conditions, a crucial function in PV systems to ensure maximum energy extraction. Physical simulations were also conducted to further validate the characteristic curves of the PV system. The findings reveal a remarkable consistency in PV module performance across all three methodologies, demonstrating their reliability in accurately assessing energy generation. The integration of experimental, ANN-based, and simulation approaches highlights their potential for improving the predictive accuracy and efficiency of PV systems, particularly under diverse operational conditions.</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

Comparative analysis of photovoltaic system characteristics and performance through physical simulation, ANN modelling and experimental investigation

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年2月3日

查看全文
  • Aschenaki Tadesse Altaye,
  • Piroska Víg,
  • István Farkas

摘要

This study comprehensively analyses photovoltaic (PV) system performance by examining its characteristic curves and conducting a comparative evaluation using three methodologies: experimental investigation, artificial neural network (ANN) modelling, and physical simulation, which constitutes the study’s novelty. Recognizing research gaps such as data limitations, incomplete models, ANN applicability, measurement inaccuracies, and the need for validated experimental inputs in PV energy harvesting, an outdoor experimental setup, integrated with LabVIEW software, was employed to collect voltage and current data from a PV module under real-world conditions. Based on these data, an ANN model was developed to predict power generation, achieving a high regression value of 0.9963, indicating a strong correlation with measured results. Furthermore, the ANN model was utilized for maximum power point tracking (MPPT) to optimise power output under varying environmental conditions, a crucial function in PV systems to ensure maximum energy extraction. Physical simulations were also conducted to further validate the characteristic curves of the PV system. The findings reveal a remarkable consistency in PV module performance across all three methodologies, demonstrating their reliability in accurately assessing energy generation. The integration of experimental, ANN-based, and simulation approaches highlights their potential for improving the predictive accuracy and efficiency of PV systems, particularly under diverse operational conditions.