Development and Performance Evaluation of a Hybrid AI-Based Method for Defects Detection in Photovoltaic Systems

Ali Thakfan and Yasser Bin Salamah ()
Additional contact information
Ali Thakfan: Joint Master’s Program in Renewable Energy, Deanship of Graduate Studies, King Saud University, Riyadh 11473, Saudi Arabia
Yasser Bin Salamah: Department of Electrical Engineering, King Saud University, Riyadh 11421, Saudi Arabia

Energies, 2025, vol. 18, issue 4, 1-19

Abstract: Maintenance and monitoring of solar photovoltaic (PV) systems are essential for enhancing reliability, extending lifespan, and maintaining efficiency. Some defects in PV cells cannot be detected through output measurements due to the string configuration of interconnected cells. Inspection methods such as thermal imaging, electroluminescence, and photoluminescence are commonly used for fault detection. Among these, thermal imaging is widely adopted for diagnosing PV modules due to its rapid procedure, affordability, and reliability in identifying defects. Similarly, current–voltage (I-V) curve analysis provides valuable insights into the electrical performance of solar cells, offering critical information on potential defects and operational inconsistencies. Different data types can be effectively managed and analyzed using artificial intelligence (AI) algorithms, enabling accurate predictions and automated processing. This paper presents the development of a machine learning algorithm utilizing transfer learning, with thermal imaging and I-V curves as dual and single inputs, to validate its effectiveness in detecting faults in PV cells at King Saud University, Riyadh. Findings demonstrate that integrating thermal images with I-V curve data significantly enhances defect detection by capturing both surface-level and performance-based information, achieving an accuracy and recall of more than 98% for both dual and single inputs. The approach reduces resource requirements while improving fault detection accuracy. With further development, this hybrid method holds the potential to provide a more comprehensive diagnostic solution, improving system performance assessments and enabling the adoption of proactive maintenance strategies, with promising prospects for large-scale solar plant implementation.

Keywords: solar PV; defects; faults; machine learning; thermal images; I-V curves; neural networks; transfer learning (search for similar items in EconPapers)
JEL-codes: Q Q0 Q4 Q40 Q41 Q42 Q43 Q47 Q48 Q49 (search for similar items in EconPapers)
Date: 2025
References: View references in EconPapers View complete reference list from CitEc
Citations:

Downloads: (external link)
https://www.mdpi.com/1996-1073/18/4/812/pdf (application/pdf)
https://www.mdpi.com/1996-1073/18/4/812/ (text/html)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:gam:jeners:v:18:y:2025:i:4:p:812-:d:1587478

Access Statistics for this article

Energies is currently edited by Ms. Agatha Cao

More articles in Energies from MDPI
Bibliographic data for series maintained by MDPI Indexing Manager ().