| Autor: |
Mateusz Wójcik, Mirosław Włuka, Andrzej Bielecki, Marzena Bielecka, Tomasz Barszcz |
| Rok vydání: |
2015 |
| Předmět: |
|
| Zdroj: |
Applied Condition Monitoring ISBN: 9783319204628 |
| DOI: |
10.1007/978-3-319-20463-5_20 |
| Popis: |
A common technique used to decrease a cost of wind turbine maintenance is a remote monitoring. Apart from the development of several advanced diagnostic methods for wind turbines there is a need to prepare an early warning tool which would work continuously in real-time and focus the attention on potentially dangerous cases. A research using the resonance neural networks made so far by the authors gave positive results. Systems based on the ART-2 networks were able to perform a classification of operational states of a horizontal axis wind turbine. In this paper the innovative idea of using the ART-2 network is applied to data from vertical axis wind turbines. The system, which were composed by ART-2 and new signal normalization procedures based on a stereographic projection, was implemented and tested. Simulations of a system operation showed that it is capable to perform an efficient state classification. |
| Databáze: |
OpenAIRE |
| Externí odkaz: |
|
