Kinematic Control of a Mecanum Mobile Robot using Time-Varying Model Predictive Control

Mecanum wheeled mobile robots have increasingly caught researchers' attention due to their wide range of applications and higher maneuverability. However, their complex structure makes their trajectory tracking and control problem more challenging. This paper presents a cascaded controller for trajectory tracking of a four-wheel Mecanum mobile robot. The cascaded controller consists of a linear time-varying model predictive controller (LTV-MPC) and an inner PID controller. The proposed controller does not require the system's dynamics model. To this end, the linear time-varying error model used to design the MPC is obtained from the robot's kinematics model. The angular velocities obtained from the predictive controller are then mapped into the required commands for the DC motors using a PID controller. The PID controller is fine-tuned considering the motor's dead zone using the Genetic Algorithm. Simulations and experimental results demonstrate the superiority of the cascaded controller over the regular LTV-MPC. The proposed algorithm can also be used for other real-time implementations of robotic tasks.