Modal parameter estimation for operational wind turbines

Copyright © Inria (2014). Wind turbines are time-varying systems excited by loads due to the wind and to the interaction between blades, tower and drivetrain. In order to predict in advance failures or damages, it is useful to estimate the modal parameters of the system and to analyze their variations in operating conditions. Since it is very difficult to measure the loads, the modal identification procedure needs to rely only on the output measurement data. Operational Modal Analysis (OMA) is well suited for the estimation of modal parameters in several cases. One of the main assumptions on which the OMA theory is based is the linear time-invariance of the system. This assumption holds in the case of parked wind turbines, but not in the case of operating wind turbines. Therefore, the OMA technique needs to be adapted in order to be applied to linear time-variant systems. Alternatively, time-variant systems could be converted to time-invariant ones before applying the classical OMA. The Multi-Blade Coordinate transformation (MBC) allows describing the dynamic interaction between the nonrotating components and the rotor. By applying the MBC, the time-periodic system is converted into a time invariant one and conventional OMA can be applied to estimate the modal parameters. In this paper, a multibody simulation model of a wind turbine is used to assess how numerical and experimental techniques can be combined for Structural Health Monitoring (SHM) of operating wind turbines.status: publishe