Application of Neural Networks for Detection of Changes in Nonlinear Systems

A nonparametric structural damage detection methodology based on nonlinear system identification approaches is presented for the health monitoring of structure-unknown systems. In its general form, the method requires no information about the topology or the nature of the physical system being monitored. The approach relies on the use of vibration measurements from a "healthy" system to train a neural network for identification purposes. Subsequently, the trained network is fed comparable vibration measurements from the same structure under different episodes of response in order to monitor the health of the structure and thereby provide a relatively sensitive indicator of changes (damage) in the underlying structure. For systems with certain topologies, the method can also furnish information about the region within which structural changes have occurred. The approach is applied to an intricate mechanical system that incorporates significant nonlinear behavior typically encountered in the applied mechanics field. The system was tested in its "virgin" state as well as in "damaged" states corresponding to different degrees of parameter changes. It is shown that the proposed method is a robust procedure and a practical tool for the detection and overall quantification of changes in nonlinear structures whose constitutive properties and topologies are not known.