Model-based Controller Design for Automatic Two Wheeled Self Balancing Robot

Abstract

Autonomous mobile robots have been employed successfully for a variety of purposes throughout the past decade. Self-balancing robots (SBR) have recently received increasing attention and demonstrated outstanding performance. Several scenarios, such as robotics weaponry, inverted pendulums, bicycles, and rocket launchers, make substantial use of the mechanics of the two-wheeled self-balancing robot (TWSBR). This structure will stay unstable if a trustworthy means of control is not implemented. The dynamic structure of TWSBR exhibits extremely nonlinear behaviour. The primary goal of this study is to stabilize the TWSBR by transferring the signal from the controller's output to the robot's chassis. The TWSBR mathematical model is first derived using Newton's principles. Second, a Model Predictive Controller (MPC) is developed and put into place to ensure that the TWSBR remains stable even in the presence of external disturbance. Finally, the performance of MPC is evaluated on TWSBR in self-balancing against that of a traditional Proportional Integral Derivative (PID) controller. The optimal controller for TWSBR control is then identified and put into practice in real-time.