Error-Model Predictive Control of Wheeled Mobile Robots for Minimum-Time Trajectory Tracking

In this paper, we propose an error-based four state kinematic model of a wheeled mobile robot with tricycle drive for trajectory tracking control. The trajectory tracking algorithm was developed for a discrete system and implemented using a model predictive control (MPC) approach. The objective function of the MPC is minimized with particle swarm optimization (PSO). The minimum-time trajectory used in the experiments satisfies velocity, acceleration, and jerk constraints, which we used to indirectly describe the dynamic properties of the wheeled mobile robot (WMR). Simulation results showed robust performance in the presence of various non-ideal conditions, such as measurement noise, delays, and constrained control velocities. Consequently, we were also able to apply the approaches from the simulations to a real robot platform to confirm the real-time applicability. Our proposed control algorithm and trajectory generation approach are well suited for automated guided vehicles (AGVs) used in logistics in industrial environments, where efficient operation depends on minimizing travel time.