Path Planning for Lunar Surface Person-Following Robot via Flatness-Based Safe-MPC With Virtual Disturbances

As lunar exploration intensifies, astronauts will encounter increasing challenges in effectively completing surface missions. The development of sophisticated assistive robots, capable of tracking and supporting astronauts to reduce their workload, has emerged as a significant research focus. However, current challenges with person-following robots include maintaining reliable and effective tracking in complex disturbance environments, managing limited computational resources, and ensuring operational safety. In this article, we focus on the integration of motion planning and high-performance person-following robots for robust tracking control of robot. Considering differentially flat system property has good performance in problems about finite-time transition between two set-points and the tracking mission needs to plan the path in the safe region, we present a flatness-based safe-model predictive control (MPC) with virtual disturbances that combines the advantages of MPC and differentially flat systems to drive the robot along a real-time and high-performance trajectory while satisfying practical constraints. By solving the optimization problem in the flat output space, we obtain the reference trajectory from nominal point to target point and design the tube size according to the mission scenario. Further, to obtain such tube, we introduce virtual disturbance into the differentially flat system, and then use TMPC to track the reference path. The simulation results in lunar scenario demonstrate that the robot can track the target well in the safe area, and the system is robust.