Vibration-based angular speed estimation for multi-stage wind turbine gearboxes

Cédric Peeters, Quentin Leclère, Jéröme Antoni, Patrick Guillaume, Jan Helsen

Onderzoeksoutput: Articlepeer review

11 Citaten (Scopus)

Samenvatting

Most processing tools based on frequency analysis of vibration signals are only applicable for stationary speed regimes. Speed variation causes the spectral content to smear, which encumbers most conventional fault detection techniques. To solve the problem of non-stationary speed conditions, the instantaneous angular speed (IAS) is estimated. Wind turbine gearboxes however are typically multi-stage gearboxes, consisting of multiple shafts, rotating at different speeds. Fitting a sensor (e.g. a tachometer) to every single stage is not always feasible. As such there is a need to estimate the IAS of every single shaft based on the vibration signals measured by the accelerometers. This paper investigates the performance of the multi-order probabilistic approach for IAS estimation on experimental case studies of wind turbines. This method takes into account the meshing orders of the gears present in the system and has the advantage that a priori it is not necessary to associate harmonics with a certain periodic mechanical event, which increases the robustness of the method. It is found that the MOPA has the potential to easily outperform standard band-pass filtering techniques for speed estimation. More knowledge of the gearbox kinematics is beneficial for the MOPA performance, but even with very little knowledge about the meshing orders, the MOPA still performs sufficiently well to compete with the standard speed estimation techniques. This observation is proven on two different data sets, both originating from vibration measurements on the gearbox housing of a wind turbine.

Originele taal-2English
Artikelnummer012053
Pagina's (van-tot)1-9
Aantal pagina's9
TijdschriftJournal of Physics: Conference Series
Volume842
Nummer van het tijdschrift1
DOI's
StatusPublished - 2 jun 2017
EvenementInternational Conference on Damage Assessment of Structures - Kitakyushu, Japan, Kitakyushu, Japan
Duur: 10 jul 201712 jul 2017
Congresnummer: 12
http://www.damas.ugent.be

Vingerafdruk

Duik in de onderzoeksthema's van 'Vibration-based angular speed estimation for multi-stage wind turbine gearboxes'. Samen vormen ze een unieke vingerafdruk.

Citeer dit