Featured in Best in class – 01-2020

A numbers game: AI for PV faults

As part of his work as guest researcher at the Fraunhofer Institute for Solar Energy (ISE), Mücahid Candan specializes in the use of artificial intelligence in photovoltaic systems. Here, he tells pv magazine about his ongoing project: deep learning-based fault detection using neural network ensemble (DEFNE).

By

Blathnaid O’Dea

03 Dec 2024

Photo: Mücahid Candan

pv magazine: Can you explain what DEFNE is?Mücahid Candan: The project focuses on detecting faults and anomalies in photovoltaic systems by analyzing individual measurement samples – collected at intervals of five to 15 minutes – typical in such systems.DEFNE is an acronym for “Deep learning-based fault detection using neural network ensemble.” It’s also a popular …

This article is part of pv+

Subscribe digitally to unlock exclusive investigative reporting, expert insights, in-depth analysis, and premium background coverage. Your subscription also includes full access to the digital edition of our magazine.