Trajectory tracking of Omni-directional Mobile Robots via Predictive Control plus a Filtered Smith Predictor

Abstract

In this work, an unconstrained Model Predictive Control (MPC) controller, and a filtered Smith predictor (FSP) structure are applied as a solution for a trajectory tracking problem of an omni-directional mobile robot through a networked based control scheme. The FSP is able to deal with the process dead-time, due to communication delay, and also mitigate measurement noise improving the robustness of the controller, while the control signal is calculated by simply applying the MPC explicit solution. The proposed controller uses a cascade structure where the innermost control loop is responsible for the velocity control, while the outer loop controls the position error. The performance of the controller is evaluated using experimental results when controlling the mobile robot in different trajectories.