System Identification of Flexible Beam Structure Using Artificial Neural Network

Abstract

This paper presents the development of a nonparametric model that represents the dynamic behaviour of a flexible beam system utilizing several artificial neuralnetwork algorithms. Input-output data used in this study isobtained from Finite Difference algorithm's simulation. Thealgorithm is validated through comparison of its natural frequencies of vibration with the theoretical values. For system identification, non-parametric approach namely ArtificialNeural Network (ANN) is utilized in this study. First is by using Multilayer Perceptron (MLP) and the second method isby using Radial Basis Function (RBF). Several validation testswere carried out to measure the performance of developed model for each technique. Results indicated a superiority for both techniques in modelling a flexible beam structure.