Testing PID and MPC Performance for Mobile Robot Local Path-Following

Abstract

This paper outlines the online performance of a control law based on PID (proportional-integral-derivative) controllers and MPC (model predictive control) for mobile robot local path-following. Both techniques share the use of a set of different dynamic models. PID controllers are used for controlling the speed of the robot's wheels, while high level algorithms compute the necessary wheel speeds in order to generate a motion that approaches the vehicle towards the desired path. Meanwhile, local MPC is implemented by computing the horizon of suitable coordinates that arise from the set of command input combinations. Therefore, command speeds that correspond to the desired point are obtained by minimizing a cost function in which the population of the available coordinates is taken into account.