Use this link to cite:
http://hdl.handle.net/2183/33568 Detección de fallos en aerogeneradores flotantes mediante redes neuronales usando OpenFAST
Loading...
Identifiers
Publication date
Authors
Galeote, Ignacio
Aimara Andrade, Bryan Alexander
Esteban, Segundo
Santos, Matilde
Advisors
Other responsabilities
Journal Title
Bibliographic citation
Galeote, I., Aimara, G.A, Esteban, S., Santos, M., 2023. Detección de fallos en aerogeneradores flotantes mediante redes neuronales usando OpenFAST. XLIV Jornadas de Automática. 144-149. https://doi.org/10.17979/spudc.9788497498609.144
Type of academic work
Academic degree
Abstract
[Resumen] Uno de los principales problemas de la energía eólica es el de la continuidad en la producción, exacerbado en el caso de la tecnología flotante dada la complejidad añadida de las circunstancias ambientales. Dada la variabilidad intrínseca del viento, que conlleva la producción irregular de energía, es de especial importancia detectar y minimizar tanto la frecuencia como la gravedad de fallos o averías en las máquinas. En el presente trabajo se ha estudiado una turbina de referencia flotante en alta mar de 5 MW y se han simulado fallas de diversos elementos estructurales mediante el software OpenFAST del NREL, mediante una técnica de acople de simulaciones. Después, se ha entrenado una red neuronal empleando MATLAB con el objetivo de identificar aquellos sensores más adecuados para detectar estas anomalías, así como su respuesta característica que permita hacer un diagnóstico rápido y fiable del fallo.
[Abstract] One of the main problems of wind energy is that of production continuity, exacerbated in the case of floating devides due to the added complexity of the environmental loads. Given the intrinsic variability of wind, which leads to irregularities in energy production, it is of particular importance to detect and minimize both the frequency and severity of machine failures or malfunctions. In this work, a 5 MW offshore floating reference turbine has been studied; and failures of various structural elements have been simulated using NREL’s OpenFAST software, using simulation coupling techniques. Then, a neural network has been trained using MATLAB, with the aim of identifying the most suitable sensors to detect these anomalies, as well as their characteristic response that allows a fast and reliable diagnosis of the failure.
[Abstract] One of the main problems of wind energy is that of production continuity, exacerbated in the case of floating devides due to the added complexity of the environmental loads. Given the intrinsic variability of wind, which leads to irregularities in energy production, it is of particular importance to detect and minimize both the frequency and severity of machine failures or malfunctions. In this work, a 5 MW offshore floating reference turbine has been studied; and failures of various structural elements have been simulated using NREL’s OpenFAST software, using simulation coupling techniques. Then, a neural network has been trained using MATLAB, with the aim of identifying the most suitable sensors to detect these anomalies, as well as their characteristic response that allows a fast and reliable diagnosis of the failure.
Description
Editor version
Rights
Atribución-NoComercial-CompartirIgual 3.0 España


