Autonomous Transportation Robots for Finished Vehicle Docking in RORO Logistics Terminal: Design, Control, and Implementation

The continuous growth of global automobile trade volume has intensified the pressure on Roll-on/Roll-off (RORO) terminals as critical transshipment hubs. Traditional manual vehicle transshipment is inefficient and has become a significant operational bottleneck. Therefore, research on autonomous transportation of finished vehicles using robotic technology has increased significantly. However, an Autonomous Transportation Robot (ATR) designed for this task requires a high degree of autonomy, particularly for accurately docking with finished vehicles to facilitate handling. In this paper, we propose an Iterative Learning Model Predictive Control (ILMPC) framework, enhanced by a Radial Basis Function (RBF) neural network, to achieve autonomous docking. This approach significantly improves the robot's operational flexibility and docking precision. To improve the data-driven model identification, this paper proposes an optimal excitation signal generation method that utilizes vehicle posture data from previous iterations. Finally, by designing three docking scenarios for physical vehicle experiments, the results show that with the increasing number of iterations, the docking accuracy of the robot can be gradually enhanced, verifying the effectiveness and feasibility of this method.