Hybrid Enhanced Control Architecture with Physics-Guided Constraints for Tiltrotor Hovering Stability

Tiltrotor aircraft exhibit complex aerodynamic coupling and interference effects, which pose significant challenges to achieving effective fuselage stability control. In this paper, an aerodynamic model is first developed for an advanced tilt-rotor aircraft configuration. For the first time, active disturbance rejection control (ADRC) theory is integrated with a physics-informed neural network (PINN) to develop a Hybrid control strategy for dynamically compensating variable errors. By constraining system uncertainties through physical principles, a real-time estimation and robust suppression mechanism for lumped disturbances is proposed, effectively reducing the traditional control's reliance on prior models. Finally, the effectiveness of this method in tilt-rotor aircraft stability control is validated through comparative experimental analysis.