Optimal Integration of Data Fusion in Solar Power Analytics: Enhancing Efficiency and Accuracy

Darío González-Cruz¹, Franky Jiménez-García¹, Javier Gamboa-Cruzado^1,2, Edward R. Luna Victoria¹, María Lima Bendezú³, Reem Attasi⁴

¹ César Vallejo University, Trujillo, Peru

² National University of San Marcos, Trujillo, Peru

³ National University Micaela Bastidas of Apurímac, Apurímac, Peru

⁴ Higher Colleges of Technology, United Arab Emirates

Emails: darioucv@ucvvirtual.edu.pe; fjimenezga@ucvvirtual.edu.pe; jgamboac@ucv.edu.pe; erubio@ucv.edu.pe; mlima@unamba.edu.pe; jgamboac@unmsm.edu.pe; ratassi@hct.ac.ae

Abstract

At the forefront of sustainable energy solutions lies renewable energy, particularly solar power. Nevertheless, the optimization of solar power systems necessitates comprehensive analytics, especially for proactive maintenance fault anticipation. This research evaluates data fusion techniques using both linear and non-linear regression models for predicting faults in solar power plants. The study begins with careful data preparation processes to ensure clean and harmonized data sets that include irradiation, temperature, historical fault records, and yield. Linear regression techniques provide insights into straightforward correlations while non-linear models go deep into complex relationships within the data. The results indicate positive outcomes demonstrating the potential of these fusion techniques as far as improving accuracy in fault prediction is concerned. These findings highlight the importance of refining data preparation prior to any fusion process and recommend further exploration into more advanced fusion methodologies. This paper helps advance proactive maintenance strategies for solar power plants thereby making this source of energy more dependable and resilient.

Keywords: Solar Energy Analytics; Information Fusion; Photovoltaic Systems; Energy Harvesting Analysis; Multi-source Data Fusion; Solar Power Optimization; Machine Learning; Performance Enhancement.