A Robust 6-DOF Dynamic Soaring Framework Incorporating Variable Aerodynamic Derivatives

Dynamic soaring is a flight technique that exploits wind shear for sustained flight. It is commonly observed in birds such as albatrosses and holds significant potential for unmanned aerial vehicle (UAV) missions. Previous research has primarily focused on trajectory generation using direct optimal control or differential flatness. This paper proposes an enhancement to the existing six-degree-of-freedom (6-DOF) trajectory generation method based on differential flatness. The proposed formulation includes sideslip and accounts for all stability and control derivatives. A Vortex Lattice Method (VLM) solver is then used to compute steady aerodynamic forces and moments, which are compared against the constant-derivative-based trajectories. To assess the validity of the constant-derivative assumption, a 6-DOF UAV model is simulated in a dynamic soaring orbit with stability augmentation provided by a Linear Quadratic Regulator (LQR). The observed divergence in this simulation highlights the limitations of the constant-derivative approach. Trajectory generation is then refined by incorporating the variation of aerodynamic derivatives with flight conditions, using data from a lookup table generated using a VLM solver. The effectiveness of this improved approach is demonstrated through simulation results. The main contributions of this work are: (i) a differential-flatness-based dynamic soaring formulation that includes sideslip and full derivative coupling, (ii) a validation framework that exposes limitations of constant-derivative assumptions, and (iii) a lookup-table-based trajectory generation method that enhances stability and realism, providing a practical pathway toward experimentally realizable dynamic soaring trajectories.