Unified Terrestrial/Aerial Motion Planning for HyTAQs via NMPC

Abstract

In this study, an NMPC-based planning approach is proposed for Hybrid Terrestrial and Aerial Quadrotors (HyTAQs) to achieve safe and efficient hybrid locomotion in unknown environments. The approach ensures bimodal dynamic feasibility and fully exploits the high efficiency and accessibility of HyTAQs. Equipped with the complementarity constraints for hybrid dynamics, the unified NMPC enables simultaneous optimization of the full-state trajectory and the inherent locomotion mode without the computational burden of mixed-integer programming. Also, the uncertainty bounds are evaluated along the nominal trajectory in a receding horizon manner and incorporated within the geometry constraints for the safety and robustness. Further, the bimodal energy efficiency is fully considered by the topology-guided path searching, achieving more efficient mode selection. Then, a compact unified terrestrial-aerial planner is implemented on the HyTAQ and validated through various comparisons and navigation tests. Experimental results verify the high efficiency and robustness of the proposed approach.