Coupling-Guided Nonlinear Model Predictive Control of an Aerial Manipulator under Constrained Direct Pitch Actuation
Year: 2026
Type: preprint
Abstract: This study investigates long-range end-effector target reaching for an aerial manipulator under constrained direct pitch actuation. A coupling-guided two-stage nonlinear model predictive control method is developed using a five-degree-of-freedom longitudinal prediction model derived from the coupled dynamics of a quadrotor and a two-joint manipulator. During the long-range approach, a coupling-guided allocation penalty and asymmetric actuator costs reshape the relative costs of feasible base–joint motion and input sequences, thereby reducing reliance on the direct pitch channel. Near the target, the control objective shifts to accurate end-effector regulation with base-drift suppression and prediction-terminal contraction. Comparative and ablation studies show that the proposed strategy reduces direct pitch-channel demand while maintaining comparable final positioning accuracy, and that the guidance term changes control-sequence selection rather than simply transferring pitch demand to the joints. Sensitivity analyses further demonstrate that the unloading behavior persists over the tested guidance-coefficient combinations and direct pitch-torque limits. The results indicate that base–manipulator dynamic coupling can provide an additional pathway for task realization under constrained attitude-control authority through coordinated multibody motion and actuator allocation.
Source: Research Square
Authors: Qingling Lin, Gefei Shi, Zhigang Wu
Institution: Sun Yat-sen University
Language: English
Subfield: Control and Systems Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: green
Award: G708414365
OpenAlex ID: W7225879550
