Machine Learning and Neural Network for Maintenance Management

A novel Non-Destructive Test (NDT) is presented in this paper. It employs a radiometric sensor that measures the infrared emissivity of the solar panel surface embedded in an unmanned aerial vehicle. The measurements provided by the sensor will determine if the panel is healthy, damaged or dirty. A...

Full description

Bibliographic Details
Authors: Gómez Muñoz, Carlos Quiterio, García Márquez, Fausto Pedro, Arcos Jiménez, Alfredo
Format: book part
Publication Date:2017
Country:España
Institution:Universidad de Castilla-La Mancha
Repository:RUIdeRA. Repositorio Institucional de la UCLM
OAI Identifier:oai:ruidera.uclm.es:10578/16084
Online Access:https://doi.org/10.1007/978-3-319-59280-0_96
http://hdl.handle.net/10578/16084
Access Level:Open access
Keyword:Fault detection
Infrared sensor
Radiometry
Solar plants
Photovoltaic panels
Fault detection and diagnosis
Description
Summary:A novel Non-Destructive Test (NDT) is presented in this paper. It employs a radiometric sensor that measures the infrared emissivity of the solar panel surface embedded in an unmanned aerial vehicle. The measurements provided by the sensor will determine if the panel is healthy, damaged or dirty. A thermographic camera has been used to check the temperature variations and validate the results by the sensor. The study shows that the amount of dirt influences the temperature on the surface and the energy generated. Similarly, faults in photovoltaic cells influence the temperature of the panel. The NDT system is less expensive than traditional thermographic sensors or cameras. Early detection of these problems, together with an optimal maintenance strategy, allows to reduce costs and increase the competitiveness of this renewable energy source.