Flight Test of L 1 Adaptive Control on 120-kg-Class Electric Vertical Take-Off and Landing Vehicles
Zian Wang,Zheng Gong,Chi Zhang,Jun He,Shengchen Mao
DOI: https://doi.org/10.1109/access.2021.3132963
IF: 3.9
2021-01-01
IEEE Access
Abstract:Electric vertical take-off and landing vehicles (eVTOLs) are a novel class of transportation that can facilitate point-to-point travel. However, the control of eVTOLs traversing from the gyro to the fixed-wing flight mode poses daunting challenges in the context of modeling, the design of the control algorithm, flight management, simulation, verification, and testing. This paper proposes the design of an $\mathcal {L} _{1}$ -augmented autopilot that is implemented on the 120-kg-class large-scale Electric Transportation 120 platform of the Commercial Aircraft Corporation of China, Ltd. Significant advances in fast rotor modeling according to blade element momentum theory, virtual flight test techniques in wind tunnels, a layered design of the architecture for flight management, and fast techniques for validation and verification speed-up the development of the flight control system (FCS). The state-of-the-art $\mathcal {L}_{1}$ adaptive control architecture combined with dynamic inversion type control allocator is particularly suitable for dealing with nonaffine control problems encountered with aerodynamic uncertainties. Implement of $\mathcal {L} _{1}$ adaptive control theory significantly reduces the parameter tuning cycle to achieve the desired closed-loop tracking performance. Real-word flight tests have confirmed the effectiveness of $\mathcal {L}_{1}$ -augmented algorithm and customized FCS.
computer science, information systems,telecommunications,engineering, electrical & electronic