Damage identification in laboratory beams using neural networks

This paper presents the application of Artificial Neural Networks (ANN) in the identification of damage in simple laboratory beam structures. The application of ANNs improves the non-destructive vibration-based damage identification method. In this method, the damage is identified on the basis of the variations of dynamic parameters without knowledge of the initial values of undamaged structures. The assessment of the state of a structure relies on the comparison of the structure natural frequencies obtained from the systems with additional masses placed in different nodes. In the presented experimental examples, ANNs are applied for the analysis of the dynamic response of beams for the location of damage and the extent of damage.