| Description |
Adel Mellit
(Faculty of Sciences and Technology, University of Jijel, Algeria) Abstract:
Photovoltaic (PV) plants may experience breakdowns or serious energy losses due to aging or damage-induced failures that occur mainly in solar PV panels. Therefore, fault detection, localization, and diagnosis are critical for increasing the efficiency and dependability of PV panels. The hotspot issue is the most hazardous and invisible defect, which cannot be observed by visual inspection. This can be identified using an infrared camera, and if the defect is not identified early, it can lead to a risk of fire. In this presentation, I discuss the development of a method for the classification of defective PV panels based on infrared images. The idea is to develop a TinyML model to classify certain invisible defects and then integrate the impulse into an edge device for real-time applications. I will present also the designed prototype. The preliminary results enabled an evaluation of the methodology's advantages and limitations to improve the performance of the TinyML model.
|
CMSP Informal Seminar: TinyML model for fault diagnosis of photovoltaic panels using infrared images
Go to day