This project is a two-wheeled self-balancing robot (TWSBR). The simulations are based in OpenModelica and MATLAB. The controllers implemented are:

• Proportional-Integral-Derivative (PID)
• Linear Quadratic Regulator (LQR)
• Nonlinear Model Predictive Control (NMPC)

Open Modelica is used for several experiments and the models of the system. The files are distributed as. In here, the PID and LQR controllers are simulated. Although, the gains are computed within MATLAB LQR script.

• SegwaySystem.mo is the document that holds all the components together and the one on which the experiments where made.
• SegwayDynamics.mo contains the dynamics of the system, you can change parameters of the model, initial states and disturbances. As input, it accepts torque and as output the state systems.
• LQRController.mo is the controller, you can set the offline gains, max absolute torque of the motors.
• NoiseMatrix.mo is the component that add a drift noise to the model.

To run the code, load the four codes into open modelica and simulate SegwaySystem.mo.

MATLAB is used to validate the dynamics of the system. At the same time, it is used to simulate the MPC controller for both stabilization and path following. The MPC has been implemented as a pitch, yaw, and linear velocity controller. It does not control the position as the desire is just to provide a path to follow but without time restrictions.

• The file twsbr_dynamics.m contains the function that describes the dynamics of the system.
• The file twsbr_lqr.m contains the simulation of the LQR controller. Just hit play.
• The file twsbr_nmpc.m contains the simulation of the NMPC controller with white noise and drift simulation. Hit play.
• The file twsbr_nmpc_global.m contains the simulation of the NMPC controller without noise or drift. It attempts to follow different paths (e.g. line, circle, sinusoidal, figure8).

Additionally, a simulation of the ground is added to check if the robot hits the ground or not. See ground_contact_event.m.