Data-Driven Adaptive Predictive Cooperative Control for Multi-Robot Aircraft Skin Measurement

This paper presents a novel approach to addressing the relative position problem in aircraft skin measurement based on data-driven predictive cooperative control (DDPCC) for multi-robot systems. Specifically, the study focuses on the interaction between two compound mobile robots, namely the tracker and scanner, and transforms the relative position problem into a formation control problem involving two robot chassis and a virtual leader. Leveraging the principles of predictive control, we introduce a control input criterion function, by incorporating the cross-tracking error between each robot and its neighbors, as well as its anticipated control input increment in the future. It turns out that the controller design and stability analysis of this DDPCC approach do not require explicit mathematical models. Through extensive simulations, we demonstrate the effectiveness and practicality of the proposed strategy.