Low-rank approximation of Hankel matrices in denoising applications for statistical damage diagnosis of wind turbine blades

Model order selection is a fundamental task in subspace identification for estimation of modal parameters, uncertainty propagation and damage diagnosis. However, the true model related low-rank structure dynamic system are generally unknown. In this paper, statistical methodology to actively select signal covariance-driven developed on basis analysis eigenvalue condition numbers output covariance Hankel matrix. It shown that highly sensitive random perturbations characterize noise subspace. The separated from by analyzing two parameters associated with number sensitivity, whose thresholds user-defined. A practical algorithm retrieve dynamics designed demonstrated running example simulated wind turbine blade benchmark. resultant framework then applied context detection medium-size blade. detectability small enhanced compared classic approaches robustness diagnosis increased reducing false alarms.