Unmanned Aerial Vehicle-Based Non Destructive Diagnostics

Abstract

The paper proposes a cloud platform for analyzing the radiometric infrared videos uploaded by drones which patrol large photovoltaic plants. Thanks to artificial vision algorithms, it does not require any human support to select and associate the framed PV modules to the corresponding ones in the topology of the photovoltaic plant. The algorithm implements an innovative diagnostic protocol, which evaluates the thermal state of the photovoltaic module, whichever the environmental conditions are. The data automatically computed and collected in a multimedia database provide the OM technicians with significant information to monitor the ageing of each module of the photovoltaic plant. The proposed platform also integrates a cloud-based software, named DISS, which provides quantitative and deeper information about the thermal behavior of the photovoltaic modules.