A numerical virtual flight framework based on active disturbance rejection control

Abstract

To address the strong coupling among aerodynamics, flight dynamics, and control in complex unsteady flow environments, this paper proposes a numerical virtual flight method based on Active Disturbance Rejection Control (ADRC). The proposed method aims to achieve integrated verification and optimization of aerodynamic characteristics and closed-loop control performance through high-fidelity computational fluid dynamics (CFD) simulations. A closed-loop numerical simulation framework is established, incorporating an unsteady CFD solver, a six-degree-of-freedom rigid-body dynamic model, and an ADRC-based control system. In the offline stage, a neural-network-based aerodynamic model is developed using historical flight-test data and CFD simulation data, with the angle of attack and control surface deflections as inputs and discrepancies in aerodynamic force and moment coefficients as outputs. This data-driven model is employed for the preliminary design and parameter tuning of the ADRC controller. To accommodate uncertainties caused by variations in aerodynamic characteristics under different operating conditions, an ADRC optimization strategy combining offline aerodynamic modeling and online parameter identification is further proposed, enabling adaptive compensation for unmodeled dynamics and external disturbances. The proposed method is validated through numerical simulations of the NRL7301 two-dimensional airfoil. Comparative results with a conventional Proportional-Integral-Derivative (PID) controller demonstrate that, under complex unsteady flow conditions, the ADRC-based approach achieves faster dynamic response, stronger disturbance rejection capability, and improved control stability. These results confirm the effectiveness and engineering applicability of the proposed numerical virtual flight method.