Refined State-Error-Based Model Predictive Control and Flight Testing of Eagle of Taihu 10: A Canard-Configured Tail-Sitter Technology Demonstrator

Abstract

As the size of tail-sitter electric vertical take-off and landing (eVTOL) aircraft increases, significant challenges have arisen due to a lower thrust-to-weight ratio, more nonlinear aerodynamics during transition, and slower control response. To support the design and implementation of the flight control system for the large-scale tail-sitter eVTOL Eagle of Taihu 100 (ET100), a one-third-scale, 13 kg subscale technology demonstrator, ET10, was developed. This article presents the control framework tailored for the ET series. In this framework, a refined state-error-based model predictive control (MPC) scheme is proposed. By reformulating the translational dynamics, the coupling between translation and rotation is explicitly captured. Aerodynamic feedforward and efficiency-aware control allocation are integrated to mitigate model uncertainties. A PX4 software-in-the-loop (SITL) platform was constructed by integrating data from computational fluid dynamics (CFDs) simulations and experiments to ensure a high-fidelity aerodynamic modeling. Flight simulations were first conducted to evaluate the improved algorithm in terms of tracking performance across various tasks. Next, flight tests of the ET10 were conducted, demonstrating the reliability and robustness of the scheme across the full flight envelope.