System identification and PID controller of flexible beam structure

Abstract

The purpose of this study is to investigate the application of Least Square (LS), Recursive Least Square (RLS), and Neural Network to estimate the identification of flexible beam structure and development a (Proportional-Integral-Derivative) PID controller for the system. The input and output data used for system identification process were obtained from experimental setup. The performance of system identification used, Lest Squares (LS), Recursive Least Squares (RLS), and Neural Network (NN) were verified using Mean Square Error (MSE) technique and correlation test. Comparative assessment was conducted to compare all the result obtained and the best transfer function was obtained from NN system identification approach with smallest MSE value of 7.9577×10−5. The transfer function has been used to help the development of control system to suppress unwanted vibration of the flexible beam system. In this study, a PID controller has been proposed to use for the vibration suppression. This controller was tuned by using heuristic tuning in Matlab SIMULINK simulation environment. However, it was found that PI controller has performed as a good controller to suppress the vibration of flexible beam system with MSE value of 7.7671×10−5.