Use of a Spatially Adaptive Thresholding Method for the Condition Monitoring of a Wind Turbine Gearbox
Abstract
Condition monitoring of wind turbine gearboxes is an important practice in order to determine the state of the wind turbine drivetrain. In this way reparative actions could be taken whenever needed, resulting in reduction of maintenance costs. In this paper, time-frequency analysis is performed on real wind turbine gearbox datasets using the empirical mode decomposition method. Then, the outlier analysis method is applied to the power of certain intrinsic mode functions of the decomposed - using the empirical mode decomposition method - gearbox experimental datasets. These intrinsic mode functions are chosen according to their frequency content. They are related to the harmonics of the meshing frequency of the damaged stage of the gearbox examined. The outlier analysis method is a well- established method in the structural health monitoring field that computes discordancy measures for data and compares them with a threshold. Here, it is used as a standard approach whose results can be used for comparison. Finally, a novel thresholding method is proposed for feature discrimination - the phase space thresholding method. It is shown that for the particular case of gear tooth damage, because of the way it manifests in the vibration signals, the phase space thresholding method proves to be a very satisfactory method that can be used for an enhanced condition monitoring strategy.
Origin : Files produced by the author(s)
Loading...