Crack and Flaw Identification in Statics and Dynamics, using Filter Algorithms and Soft Computing

Stavroulakis, Georgios Ε.; Engelhardt, Marek; Antes, H.

doi:10.1007/3-211-38134-1_5

Cited by 1 publication

(2 citation statements)

References 44 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Previous works on the solution of comparable problems use relative methodology of neural networks, however, on simpler academic applications. Namely, [2] studied an orthogonal plate with one crack in several positions. Thus, more examples, which in turn means better training, were obtained.…”

Section: Discussionmentioning

confidence: 99%

“…The location of installation is the Mount Panachaiko, Peloponnese, Greece. In order to assess the integrity of the blade structure, a damage identification algorithm based on neural networks is considered [2][3]. The proposed method belongs to the modern physics-based, data-driven methods for treating complex direct and inverse problems in mechanics.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Identification of Damage in a Wind Turbine Blade Using Mechanical Measurements and Artificial Neural Networks

Koutsianitis,

Paterakis,

E. Stavroulakis

2024

JDSIS

View full text Add to dashboard Cite

Due to the stochastic nature of environmental loadings, a lot of interest is paid in the discovery of possible damages of the involved equipment in modern industry. In wind turbines’ blades, the development of a smart structural health monitoring system is essential. In this paper, a large-scale composite wind turbine blade model is designed and used for the detection of several damage scenarios. The process is mainly based on the development of monitoring techniques which exploit the capabilities of artificial neural networks. These techniques can provide the exact position of possible damages, under given external loading scenarios. Moreover, the use of such methods decreases significantly the need of external intervention and at the same time it increases the accuracy of the whole approach. The above processes are simulated using the finite element method. The goal is to develop a neural network which realizes the correlation of measurements with damage patterns. The goal is focused on the solution of inverse problems involving elastically deformable structures, based on remote mechanical measurements. The correlation between measurements and damages, which is much more complicated in comparison to image analysis, is studied by means of neural networks.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%