Posters
Come meet the poster presenters to ask them questions and discuss their work
We would like to invite you to come and see the posters at our upcoming conference. The posters will showcase a diverse range of research topics, and will give delegates an opportunity to engage with the authors and learn more about their work. Whether you are a seasoned researcher or simply curious about the latest developments in your field, we believe that the posters will offer something of interest to everyone. So please join us at the conference and take advantage of this opportunity to learn and engage with your peers in industry and the academic community.
On 9 April at 17:15, we’ll also hold the main poster session and distinguish the 7 best posters of this year’s edition with our traditional Poster Awards Ceremony. Join us at the poster area to cheer and meet the laureates, and enjoy some drinks with all poster presenters!
We look forward to seeing you there!
PO014: Defect pattern detection in a group of wind turbines gearboxes using vibrations and deep learning
Emerson Lima, Head of Condition Analysis Center (CAC), Aqtech
Abstract
Wind Turbines Generators (WTG) are subjected to extreme environmental conditions causing them to fail before their design life, with failures predominantly happening in gearbox bearings. Failure anticipation in WTGs is of great interest because gearbox components can be replaced up tower preventively, at a lower cost, avoiding possibly gearbox replacement with higher cost and intervention time. In this sense, vibration monitoring has been widely used, since, according to ISO 16079-2, it is one of the most sensitive symptoms of WTGs components degradation. It is known that due to the characteristics of the wind, the power generated by the turbine and the speed of the rotating components vary continuously. This behavior generates dispersion and mixing of the frequency components in the vibration spectrum. The vibration signature of bearings is considered as non-stationary, in addition to having shaft and gear frequencies components, being subject to noise and variations in speed and load. Due to the complexity of this signature, the fault signature can be masked, making its detection more difficult, especially in early stages, requiring, therefore, robust methodologies. ISO 10816-21 provides limits for RMS vibration of WTG components like main bearing, gearbox, and generator, making it possible to set alarm levels. However, in some cases the development of bearing failures does not reach sufficient RMS vibration for alarm actuation, as in the case of cage failures, or reaches, but only at advanced failure stages. Based on this, the use of intelligent search methodologies from machine learning and deep learning, for detecting anomalous vibration spectra that differ from the behavior of healthy bearings has been an efficient practice to anticipate failures.
No recording available for this poster.
