Damage Detection Formulation using Inverse Frequency Analysis incorporating Artificial Neural Network for Kirchhoff Plate Theory

Abstract

Nowadays, several different structural damage detection techniques are being developed with the goal of monitoring structure stability with high accuracy and low cost. One of the well-known techniques is inverse analysis based on model updating methods. However, the main challenges in this technique is the development of algorithms that assist in the processing of the enormous amounts of data for the inverse process. To overcome this, the Artificial Neural Network (ANN) has been used by many researchers to complement existing approaches. The integration of model updating methods and ANN requires not only a wealth of knowledge and experience in structural damage detection, but also appropriate numerical techniques, and proficiency in scripting programming languages. In this paper, the objective is to construct the formulation of structural damage detection using inverse analysis incorporating Artificial Neural Network (ANN) for Kirchhoff plate theory and to establish the source code. The output from the process is stiffness reduction ratio (SRF) while natural frequencies and mode shape as input data. Finite element method (FEM) was used in generating the formulation. The source code of the formulation has been written step-bystep and kept as simple as possible in Matrix Laboratory (Matlab) programming language. The performance of the formulation is verified against numerical work based on simulated damaged. The presented result shows that, this formulation exhibit excellent performance thus highly potential for damage detection of the plate structure.