Co-Optimization of Motion and Energy Domain for Hydrogen-Powered Hybrid UAVs: A Bi-Directional Coupling Architecture

Hydrogen-powered hybrid unmanned aerial vehicles (UAVs) are gaining increasing popularity in long-endurance aerial missions, owing to their superior energy density. Motion planning and energy management (power allocation) are two key algorithms supporting energy-efficient autonomous flight of hybrid UAVs. Most existing research concentrates on a sequential ‘planning-then-management’ architecture, in which motion and energy optimization are decoupled. In this paper, we propose a bidirectional coupling architecture to achieve cooptimization between motion domain and energy domain. In the motion domain, by incorporating real-time fuel cell efficiency from the energy domain, motion re-optimization is carried out to rescale the pre-generated polynomial trajectory into an energy-efficient motion trajectory, which is then followed by a Differential-Flatness-Based Controller (DFBC). In the energy domain, based on a motion-guided state-of-charge (SOC) reference, an adaptive equivalent consumption minimization strategy (A-ECMS) is carried out to offer optimized power allocation for the hydrogen-powered hybrid system. Finally, the proposed bi-directional coupling architecture is validated in a simulated case study on industrial plant inspection, demonstrating that the energy efficiency can be improved by over 13% as a result of the cross-domain co-optimization.