NullSpace and AutoRegressive damage detection: a comparative study

Abstract

In this paper we present the application of two damage detection methods to a laboratory tower. The first method is based on subspace identification. The second one is based on AutoRegressive modeling of the signals involved. Both methods are tested in a tower demonstrator simulating a wind turbine. They are able to correctly detect damage in the structure that is simulated by loosening some of the bolts in the joints. The results show that the first method is computationally more efficient, but the results are more stable with the second method.