AI-based Solar Panel Fault Prediction and Anomaly Detection in Ireland

Diang'a, Lathifa Jaffer (2025) AI-based Solar Panel Fault Prediction and Anomaly Detection in Ireland. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (1MB) | Preview
Preview	PDF (Configuration Manual) Download (973kB) | Preview

Abstract

The transition to renewable energy systems weather sensitive countries such as Ireland necessitates intelligent fault management and reliable solar infrastructure systems. In this research a data-driven, AI enhanced approach for anomaly detection and proactive fault prediction in solar panel systems under the unpredictable weather condition in Ireland are presented. Using the CRISP-DM framework, four datasets (Met Éireann (weather data), PVWatts (energy output data), Portugal Solar Panels (sensor data), and Dublin Airport (irradiance and climate data)) are independently analysed. Exploratory Data Analysis (EDA), LazyPredict regression benchmarking, and Isolation Forest-based anomaly detection are enacted. Top performing models, including NuSVR and GradientBoosting, attain RMSE scores as low as 0.02, and R² scores as high as 0.99 – this demonstrates robust predictive accuracy for irradiance patterns and solar output. Proactive maintenance approaches are supported by anomalies that flag 15-25% of potential faults. This study promotes a region-specific and scalable methodology that optimizes solar panel operational uptime; while also supporting Ireland’s national renewable energy targets, it also proposes a roadmap for integrating multiple datasets, addressing overfitting risks and future field deployments strategies that would strengthen real-world applicability.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Menghwar, Teerath Kumar UNSPECIFIED
Uncontrolled Keywords:	Anomaly Detection; CRISP-DM; LazyPredict; Machine Learning; Solar Panel Fault Detection; Renewable Energy
Subjects:	Q Science > QH Natural history > QH301 Biology > Methods of research. Technique. Experimental biology > Data processing. Bioinformatics > Artificial intelligence Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Artificial intelligence Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Machine learning H Social Sciences > HC Economic History and Conditions > Natural resources
Divisions:	School of Computing > Master of Science in Data Analytics
Depositing User:	Ciara O'Brien
Date Deposited:	01 Jul 2026 10:08
Last Modified:	01 Jul 2026 10:08
URI:	https://norma.ncirl.ie/id/eprint/9425

Actions (login required)

View Item