Parametric system identification and active vibration control of vibrational structures using genetic algorithm

A vibration system has advantages and disadvantages for us. Some of the disadvantages of the vibration system are discomfort, noise, malfunctioning, wear, fatigue and even destruction. An example of structure that leads to high vibration when subjected to disturbance forces is flexible plate structure. The aim of this research is to develop an Auto Regressive with eXogenous Input (ARX) model characterizing the dynamic behaviour of a two-dimensional (2D) flexible plate structure and the development of active vibration control (AVC) strategies for the structures. In order to construct the model, several sets of vibration data were obtained from the simulation of the flexible plate structures based on finite difference method. The sets of data obtained were utilised to develop ARX model using Least Squares (LS), Recursive Least Squares (RLS) and Genetic Algorithm (GA) methods. The models were validated using one step ahead (OSA) prediction, mean squared error (MSE) and correlation tests. Then, single-input single-output active vibration control (SISO-AVC) was devised using thus developed RLS and GA models. The performance of these systems was assessed in terms of comparison between uncontrolled signals, RLS-AVC and GA-AVC controlled signals in time domain, spectral density and attenuation of the signals in decibel (dB). The results show that GA is the best method in system modeling and vibration control of the simulated 2D flexible plate structures compared to RLS and LS.