Nonlinear Two-Wheeled Self-Balancing Robot Control Using LQR and LQG Controllers

Abstract

This paper studies the control of unstable and nonlinear two-wheeled self-balancing robot (TW-SB) using linear optimal control methods like LQR and LQG. The control purpose is to keep system stability around an equilibrium point while the robot is moving towards the desired point. A linearized mathematical model is described to obtain a linearized state space model in order to develop a controller and estimator. The designed controller has implemented on the nonlinear dynamics model of the robot. An optimal linear quadratic estimator has been designed to estimate the states of the system and has been connected to an optimal full state feedback regulator. The simulation results obtained in Matlab/Simulink show that the proposed controller is able to achieve the self-balancing while the wheels rotate towards the set-point position. Besides that, the controller has robustness to reject disturbances and noises applied to the system.