Guiding vector field-based guidance under wind disturbances applied to a tailsitter UAV

TL;DR

This work implements a GVF-based guidance law for tailsitter UAVs and integrates it with a state-of-the-art acceleration and attitude controller, enabling a direct comparison with trajectory-tracking guidance under wind disturbances. It extends the differential flatness framework to incorporate wind information and proves exponential stability for a single-integrator variant of the GVF, ensuring predictable tuning. Simulation results show that, in nominal agile maneuvers with small initial path deviations, GVF and trajectory-tracking perform similarly, but GVF provides smoother convergence when large deviations occur and in windy environments. The findings support the practical viability of GVF-based path-following for tailsitters and highlight wind-aware flatness as a valuable tool for robust, feedforward compensation.

Abstract

This paper develops a guidance control law based on a parametric Guiding Vector Field (GVF) and integrates it with a state-of-the-art acceleration and attitude control architecture for tailsitters. The resulting framework enables a direct comparison between traditional trajectory-tracking guidance and GVF-based path-following guidance using a realistic tailsitter model operating under windy conditions. Through extensive simulations, it is shown that for agile flight scenarios with wind and small initial position error, both guidance strategies achieve comparable tracking performance, indicating that the additional complexity introduced by the GVF formulation is not always justified. However, the GVF-based approach exhibits an advantage when initial deviation from the path is present, yielding smooth and well-behaved convergence toward the desired path. Two additional contributions support this evaluation. First, a modification of the parametric GVF is proposed that guarantees exponential stability of the tracking error dynamics for a single integrator system. Second, the differential flatness transform of a tailsitter vehicle is extended to account for explicit knowledge of the wind velocity vector.

Figures (9)

• Figure 1: The Cyclone tailsitter . Body reference frame and positive actuator input conventions.
• Figure 2: Wind velocity experienced by the as it traverses a frozen turbulence field at speed $s_r$, shown for a southwest wind with $\|\boldsymbol{v}_{w,s} \| = 10~\mathrm{m/s}$ in the frame.
• Figure 3: Reference speed $s_r$ with a final value of $V_s = 25~\mathrm{m/s}$.
• Figure 4: Tracking error norm on the circular curve. Zero-wind and $\| \boldsymbol{v}_{w,s}\|~=10~\textrm{m/s}$ case.
• Figure 5: Tracking error norm on the Lissajous curve. Zero-wind and $\| \boldsymbol{v}_{w,s}\|~=10~\textrm{m/s}$ case.