A Method of Fuzzy Algorithm in Controlling Ball and Beam through Simulation and Experiment

Abstract

In this paper, we present a way to create fuzzy controller from LQR method by using ANFIS toolbox of Matlab. First, after proving the ability of stability of this SIMO system under LQR method in Matlab/Simulink, we create a fuzzy controller through ANFIS toolbox of Matlab. The data, which is used to train, is collected from responses of system under LQR controller. Also, we present a hardware platform of ball and beam system. Under this fuzzy controller, control quality of ball and beam is better than under LQR controller in both simulation and experiment.