|
Title
|
|
|
|
Hyperspectral imaging for automated inspection of offshore wind infrastructure
| |
|
Author
|
|
|
|
| |
|
Abstract
|
|
|
|
| Wind energy will play a vital role in the transition to a carbon-neutral world. When comparing offshore to onshore wind energy, offshore wind energy has the advantage of more consistent and higher average wind speeds. This advantage over onshore wind turbines is offset by higher installation and maintenance costs of the turbines. Included in the operational and maintenance costs of the turbine is corrosion inspection. Currently, this is a dangerous, expensive manual inspection and where the result of this inspection is dependent on the experience and training of the inspector. To further reduce the O&M cost of the offshore wind turbine, a remotely operated inspection system would be beneficial for the safety, cost and objectivity of the inspection. This remotely operated platform has to include a corrosion sensor that is able to inspect the entire turbine in a timely fashion. However, current corrosion sensors have several limitations that make the in situ implementation time-consuming, costly and practically unattainable. Therefore, camera based imaging can offer a low-cost solution for corrosion detection. Normal RGB cameras have the limitation that it is challenging to distinguish corrosion defects in visually ambiguous scenarios, and thus the identification accuracy cannot be guaranteed. Hyperspectral imaging offers a balance between a chemically oriented approach and a visual inspection. Unlike RGB cameras, hyperspectral cameras capture hundreds of narrow wavelengths, generating a spectrum for each pixel. This spectral data enables differentiation between chemical compositions that are not distinguishable using broader and limited wavelength cameras. To evaluate if hyperspectral imaging is suitable for the inspection of offshore turbines, several experiments are performed using corrosion samples made in the laboratory, using accelerated corrosion tests. The thesis is split up into two parts: Technology and Applications. In the first part, Technology, we evaluate the most common corrosion inspection methods to see if they are suitable for offshore corrosion identification. In the second part, Applications, six different experiments are presented that show potential use cases of the use of hyperspectral imaging in offshore wind turbines. Four of these experiments use mineral identification to characterize corrosion on carbon steel or zinc- electroplated steel. The other two experiments use hyperspectral imaging for coating inspection during manufacturing. For each of the experiments mentioned, we show that the use of a hyperspectral camera in the short wave infrared region can be an added value to the inspection process, both during the manufacturing of the turbine and during operation. |
| |
|
Language
|
|
|
|
English
| |
|
Publication
|
|
|
|
Antwerp
:
University of Antwerp, Faculty of Applied Engineering
,
2023
| |
|
Volume/pages
|
|
|
|
x, 128 p.
| |
|
Note
|
|
|
|
:
Vanlanduit, S. [Supervisor]
:
Scheunders, P. [Supervisor]
| |
|
Full text (open access)
|
|
|
|
| |
|