A Dual-Layer Control System for Three-Vehicle Collaborative Handling Based on MPC

To address the challenges of low efficiency in large equipment transportation within confined indoor environments, as well as the insufficient payload capacity and poor maneuverability of traditional heavy-duty Automated Guided Vehicles (AGVs), A dual-layer collaborative control system based on Model Predictive Control (MPC) for three-vehicle collaborative transportation is proposed. By establishing a kinematic model for three-vehicle collaboration, The equipment's translation and central rotation modes are analyzed. A hierarchical control architecture is designed: the lower layer dynamically adjusts AGV wheel speeds via a differential drive model to automatically correct pose angles, while the upper layer adopts MPC to optimize vehicle velocities in real time, incorporating trajectory tracking errors and control input constraints to generate optimal collaborative commands for high-precision trajectory tracking. Experimental and simulation results demonstrate the effectiveness of the dual-layer control algorithm in mitigating motion deviations caused by velocity asynchrony among vehicles, achieving superior trajectory tracking performance.