Leveraging Aerodynamic Modeling for Accurate Path Tracking in Flapping-Wing Aerial Vehicles

Accurate path tracking remains a major challenge for flapping-wing aerial vehicles (FWAVs) due to complex aerodynamic coupling. Here, we propose a data-driven hierarchical control framework for underactuated FWAV. A pigeon-inspired FWAV was designed using avian scaling laws, which determine the kinematic and morphological parameters of this robot. We construct an aerodynamic model for FWAV based on a multirigid-body framework and quasi-steady blade-element theory. This model can characterize the time-varying lift and thrust within each wingbeat under a quasi-steady assumption. Furthermore, we develop a two-stage Bayesian optimization method to identify lift and thrust coefficients from wind tunnel data under various conditions, achieving a force prediction error within 5%. By introducing the above model as aerodynamic feedforward, we designed a cascaded controller capable of regulating the altitude and lateral heading motion of the FWAV. Outdoor experiments show that our proposed frameworks can accurately track a circular path, with root-mean-square errors (RMSE) of 0.4 m in radius and 0.3 m in altitude, and with path-tracking errors typically within 0.6 m. These results demonstrate the effectiveness of the proposed aerodynamic modeling and hierarchical control approach for precise, robust path following in FWAVs.