Trajectory Tracking Control of a Mobile Robot with Model Predictive Controller and Observer

This paper proposes the use of a model predictive controller with state observer to solve the trajectory tracking and kinematic model identification problem in mobile robotics field. Since it can be shown that observer and model predictive controller dynamics are independent from each other; model predictive controller based on a local linear-time-invariant model is designed by minimizing a finite horizon quadratic cost function and observer dynamics is designed by using a pole-placement technique. Performance of the proposed model predictive controller and observer pair was compared against that of a regular model predictive controller and it has been shown that the proposed controller-observer pair has better trajectory tracking performance.