Explicit Model Predictive Control for trajectory tracking with mobile robots

Abstract

In this paper the Model Predictive Control (MPC) strategy is used to solve the mobile robot trajectory tracking problem, where controller must ensure that robot follows pre-calculated trajectory. The so-called explicit optimal controller design and implementation are described. The MPC solution is calculated off-line and expressed as a piecewise affine function of the current state of a mobile robot. A linearized kinematic model of a differential drive mobile robot is used for the controller design purpose. The optimal controller, which has a form of a look-up table, is tested in simulation and experimentally.