Mode Shape based Damage Identification in Plate Using Support Vector Machine

Abstract

Monitoring of health of a structure has become most important and crucial task in many engineering field.  Recently, the algorithms like artificial neural network (ANN) and support vector machine (SVM) based on pattern recognition background gaining more popularity in the field of structural health monitoring (SHM). In the present research article SVM based technique is proposed in order to locate the damage in aluminum plate structure using raw mode shape data. Two boundary conditions namely, cantilever and fixed-free-fixed-free have been considered due to simplicity in replicating the experiments.  The damage is introduced by creating a long narrow cut (slit) of specific dimension. The proposed technique is first applied on simulated mode shape data and then it has been validated using experimental mode shape data. Finite element simulations have been carried out using Abaqus® package. Experimental set up has been developed in laboratory and laser doppler vibrometer (LDV) is used to extract the experimental mode shapes. Based on results obtained using simulation and experimental mode shape data it can be deduce that the SVM is a promising tool in the field of SHM