Wind Turbine Surface Damage Detection by Deep Learning Aided Drone Inspection Analysis
Autor: | Xiao Chen, Rasmus Reinhold Paulsen, Anders Christensen, Vladimir Fedorov, Kim Branner, Nicolai Andre Brogaard Riis, Anders Bjorholm Dahl, Asm Shihavuddin |
---|---|
Jazyk: | angličtina |
Rok vydání: | 2019 |
Předmět: |
Control and Optimization
Turbine blade Rotor blade Computer science 020209 energy Convolutional neural network (CNN) Energy Engineering and Power Technology 02 engineering and technology rotor blade Turbine lcsh:Technology law.invention damage detection drone inspection wind turbine law 0202 electrical engineering electronic engineering information engineering wind energy Electrical and Electronic Engineering Wind energy Engineering (miscellaneous) Wind power Renewable Energy Sustainability and the Environment business.industry lcsh:T Deep learning 020208 electrical & electronic engineering Training (meteorology) deep learning Damage detection Convolutional Neural Network (CNN) Drone Reliability engineering Damages Artificial intelligence Drone inspection business Wind turbine Energy (miscellaneous) |
Zdroj: | Energies, Vol 12, Iss 4, p 676 (2019) Energies; Volume 12; Issue 4; Pages: 676 Shihavuddin, ASM, Chen, X, Fedorov, V, Christensen, A N, Riis, N A B, Branner, K, Dahl, A B & Paulsen, R R 2019, ' Wind Turbine Surface Damage Detection by Deep Learning Aided Drone Inspection Analysis ', Energies, vol. 12, no. 4, 676 . https://doi.org/10.3390/en12040676 |
ISSN: | 1996-1073 |
Popis: | Timely detection of surface damages on wind turbine blades is imperative for minimizing downtime and avoiding possible catastrophic structural failures. With recent advances in drone technology, a large number of high-resolution images of wind turbines are routinely acquired and subsequently analyzed by experts to identify imminent damages. Automated analysis of these inspection images with the help of machine learning algorithms can reduce the inspection cost. In this work, we develop a deep learning-based automated damage suggestion system for subsequent analysis of drone inspection images. Experimental results demonstrate that the proposed approach can achieve almost human-level precision in terms of suggested damage location and types on wind turbine blades. We further demonstrate that for relatively small training sets, advanced data augmentation during deep learning training can better generalize the trained model, providing a significant gain in precision. |
Databáze: | OpenAIRE |
Externí odkaz: |