Key points are not available for this paper at this time.
The authors explore the application of the normalised bispectrum or bicoherence to the problem of condition monitoring of wind turbine blades. Background information is provided on this type of condition monitoring, how it differs from more conventional condition monitoring of turbo machinery, and the motivation for selecting bicoherence. Bicoherence is defined and compared with the power spectral density. Complications in collecting suitable data, and estimating the bicoherence from that data are investigated; including the requirements of very long stationary data sets for consistent estimates, and computational difficulties in handling such large data sets. Bicoherence is then applied to electrical power output data obtained from a 45 kW wind turbine. The turbine is operated in three configurations to represent normal and fault conditions. A blade with less flapwise stiffness but identical outer dimensions to the matched set of blades was fitted to simulate a damaged blade. Comparison of the results from the power spectral density and bicoherence indicates how the bicoherence might be employed for condition monitoring purposes. Slices of the bicoherence with one frequency fixed at the rate of rotation show clear differences between the configurations and substantially reduce the computational effort required to calculate the estimate.
Jeffries et al. (Thu,) studied this question.