Abstract:
This research presents the design of a robust ∞ H (Robust H Infinity Control) and LQR controller to balance the robot without falling and studies their the performance when disturbance signal added to the system.
The study begins with studying the dynamics of the mobile robot in order to create the robot model by MATLAB/Simulink. The robust ∞H and LQR controller are designed with the mass and the height of the robot as 60 kilograms and 1.7 meters respectively, and the disturbance added to the system consist of the change of the internal parameters of the robot model as mass and height of the robot and the measurement error of gyro scope, accelerometer and encoder. In the simulation experiments, they can be divided as: 1) the simulation of the robot with the initial tilt angle as 0.05 0.1, 0.18 and 0.235 radians, 2) the simulation of the robot with the height as 1.5, 1.6, 1.7 and 1.8 meters and 3) the simulation of the robot with the mass as 50, 60, 70 and 80 kilograms. From the simulation results show that the height variable can be neglect in the actual experiment due to it is very little affected. Therefore there are only two actual experiments 1) the experiments on the initial tilt angel as 0.1 and 0.2 radians and 2) the experiments with the mass as 40, 50 and 60 kilograms.
The results of the simulation and the actual experiments of the robot are consistent. The robust ∞H controller and LQR controller can stabilize when the initial tilt angle of the robot is between +0.2 radian. When there is the changing of the mass and the height of the robot, the robust ∞ H controller can maintain the stability and eliminate disturbance better than the LQR controller.