Non-linear Model Predictive Controller for Trajectory Tracking of an Omni-directional Robot Using a Simplified Model

In this paper we propose a non-linear model predictive controller architecture (NMPC) for trajectory tracking of an omni-directional mobile robot. The controller employs a simplified process model to predict the evolution of the state of the robot which allows for real-time minimization of the cost function using gradient descent methods. The cost function is chosen so as to penalize the position and orientation errors, as well as the variation of the control effort. Simulation tests are performed under different controller settings in order to evaluate the performance of system. The results are presented at the end of the paper.