Abstract
Selecting model fidelity for UAV transport of cable-suspended payloads partially filled with liquid is difficult because cable deformation, payload rotational inertia, and internal liquid motion do not affect all responses in the same way. This study examines a 3×3 cable and payload hierarchy under identical prescribed UAV support motion. The cable models separate deformable mechanics from spatial resolution within the rigid cable family. The payload models use a rigid surrogate with matched equilibrium properties to separate retained first-mode liquid mobility from intrinsic payload rotational inertia. For the adopted partially filled condition, removing rotational inertia produces a maximum payload center discrepancy more than two orders of magnitude larger than removing the retained liquid mode. The two reduction axes are also not generally separable after coupling. At a representative point with high forcing frequency, the interaction ratio reaches 40.15% for geometric response and 20.40% for required UAV thrust, while remaining 0.384% for payload center translation. Strong cancellation between successive reductions makes direct discrepancy nonmonotonic. Model admissibility is therefore evaluated over all higher fidelity ancestors before Pareto selection. Predictions frozen at withheld frequencies reproduce all evaluated Pareto sets for a normalized discrepancy limit of 0.10 at both valid blind frequencies. The continuous fidelity maps are supported for nondimensional forcing frequencies from 0.55 to 1.95. Within the adopted model family, minimum sufficient fidelity depends on the coupled prediction task rather than on subsystem complexity alone.

This publication has 26 references indexed in Scilit: