Implementation and controller design of a mobile inverted pendulum

Abstract

In this thesis we find nonlinear dynamic equations of a Mobile Inverted Pendulum (MIP) based on the inverted pendulum model. Then we linearize the nonlinear equations to obtain the linear model which was used in the controller design. Our goal is to desing two controllers, the first one is a nonlinear controller and the second one is a state feedback controller. These controllers have ability for balancing and MIP that can guarantee stability, performance and robustness for the nonlinear plant. The first controller is gain scheduling controller which is based on Linear Parameter-Varying (LPV) synthesis techniques and Linear Fractional Transformation (LFT) approaches. This controller is implemented on Matlab and connected to the robot via RS232 port. The second controller is a linear state-feedback LQR and Pole-placement controller. It is proposed to balance the MIP around equilibrium point and it is implemented on a DSPIC so that the robot can balance itself autonomously. Finally, we construct an MIP that could move forward, backward, leftward or rightward via a remote control.