A review of artificial intelligence and image processing approaches to fault detectionin photo voltaic systems
|
Full Text |
Pdf
|
|
Author |
Jovencio V. Merin
|
|
e-ISSN |
1819-6608 |
|
On Pages
|
1774-1784
|
|
Volume No. |
20
|
|
Issue No. |
20
|
|
Issue Date |
January 20, 2026
|
|
DOI |
https://doi.org/10.59018/1025200
|
|
Keywords |
corrosion, model, steel.
|
Abstract
The challenge for the growing demand of photovoltaic (PV) systems is that the prevalent performance is stemming from various PV system faults. These faults, classified as electrical, physical, and environmental, significantly reduce efficiency and system longevity. Conventional fault detection methods face limitations such as high cost, limited scalability, and environmental sensitivity. This paper reviews the advancements in artificial intelligence (AI) and image processing approaches for PV system fault detection, which offer accurate, automated, and scalability diagnoses. The study systematically investigates current trends and compares these technologies based on performance metrics, specifically, accuracy, precision, recall, and F1-score, while utilizing reputable engineering journal databases. The review covers classical image processing techniques. Classical AI/Machine Learning methods, deep learning approaches, and various hybrid methodologies that integrate various computational techniques for enhanced fault diagnosis. The synthesis highlights the capabilities of these advanced methods in achieving high detection accuracies, addressing the complex challenges of PV fault detection.
Back