Damage Detection in a Prestressed Membrane Using a Wavelet-Based Neurofuzzy System

Abstract

A neurofuzzy system has been proposed to detect damage in a structure with multiple damage sites. A prestressed square membrane representing a gossamer structure has been taken as the example structure. Damage at various locations on the structure has been modeled as reduced stiffness. The structure undergoes transverse vibration under the action of a transverse dynamic pressure. Using the finite element package, “ABAQUS,” the strain histories at different sensor locations in the structure under the action of the transverse dynamic pressure were calculated. The vibration response of the structure has been decomposed into a set of subsignals using wavelet analysis to define a damage feature index vector, which has been used as an indicator of damage in the membrane. The neurofuzzy system accepts the wavelet-based damage feature index vector as the input and gives the damage status of the structure as the output.