Automated Anomaly Detection and Localization in Solar Panels Using Deep Learning Models

Usama, Muhammad (2024) Automated Anomaly Detection and Localization in Solar Panels Using Deep Learning Models. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (4MB) | Preview
Preview	PDF (Configuration Manual) Download (1MB) | Preview

Abstract

Fault detection and localization in photovoltaic (PV) panels is a critical step in ensuring the efficiency and reliability of solar power systems. This research focuses on automating the process by applying advanced deep learning techniques, combining anomaly detection and fault localization in a comprehensive two-tier architecture. The system focuses on three key components: detecting anomalies, localizing faults, and evaluating the performance of the methodologies. For anomaly detection, the VGG19 model with transfer learning and fine-tuning was applied, achieving a training accuracy of more than 95% and a validation accuracy of more than 83%. For fault localization, the YOLOv8s model was used, which effectively localizes the faulty areas in the PV panels. The performance of these models was evaluated based on precision, accuracy, and robustness in real-world scenarios. The results show the system’s ability to significantly reduce manual inspection efforts, enhance fault detection accuracy, and improve the operational efficiency of PV installations. This research shows the potential of integrating classification and localization techniques in building intelligent monitoring systems for solar energy applications.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Simiscuka, Anderson UNSPECIFIED
Uncontrolled Keywords:	Anomaly Detection; Fault Localization; VGG19; YOLOv8s; Photovoltaic Panels; Deep Learning; Transfer Learning; Solar Energy Monitoring
Subjects:	G Geography. Anthropology. Recreation > GE Environmental Sciences Q Science > QA Mathematics > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science H Social Sciences > HD Industries. Land use. Labor > Specific Industries > Energy industries Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Machine learning
Divisions:	School of Computing > Master of Science in Artificial Intelligence
Depositing User:	Ciara O'Brien
Date Deposited:	20 Jun 2025 10:59
Last Modified:	20 Jun 2025 10:59
URI:	https://norma.ncirl.ie/id/eprint/7971

Actions (login required)

View Item