Damage Detection in a Wind Turbine Blade Based on Time Series Methods - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Conference Papers Year : 2014

Damage Detection in a Wind Turbine Blade Based on Time Series Methods


The interest in renewable energy in the European Union has increased in the past years, thus efficient energy harvesting becomes more important. For the sector of wind energy, the consequences are growing sizes of wind turbines (WTs) and erections in remote places, such as off-shore. The resulting increase of operation and maintenance costs can be counteracted by structural health monitoring systems. Different methods have been developed for detection of damages in WT blades. However, the majority are not suitable for in-service measurements or require dense sensor arrays. This paper presents a damage detection method based on autocorrelations of response accelerations. The damage sensitive feature (DSF) is developed as the Mahalanobis distance between a baseline and a current vector of the autocorrelation coefficients. Firstly, the usefulness of the DSF is assessed by using the Bayes error rate. Secondly, statistical hypothesis testing is utilized for a decision about the structural state. The procedure is applied to numerical simulations of a single WT blade with a disbonding damage scenario. The time series of accelerations are obtained from transient simulations with a simplified aerodynamic loading. The damage detection results show to be sensitive for the chosen damage scenario and are promising for prospective developments of damage detection methods in WTs.
Fichier principal
Vignette du fichier
0014.pdf (335.45 Ko) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-01020344 , version 1 (08-07-2014)


  • HAL Id : hal-01020344 , version 1


Simon Hoell, Piotr Omenzetter. Damage Detection in a Wind Turbine Blade Based on Time Series Methods. EWSHM - 7th European Workshop on Structural Health Monitoring, IFFSTTAR, Inria, Université de Nantes, Jul 2014, Nantes, France. ⟨hal-01020344⟩
170 View
360 Download


Gmail Facebook X LinkedIn More