System identification & attitude control of avian-type flyer with flight test

Abstract

This paper investigates system identification of the longitudinal and lateral linear models and control of an avian-type flyer with flight test. Longitudinal and lateral linear models are separately adopted by referring to fixed-wing aircraft cases. System identification is performed by exciting the avian-type flyer with doublet and multi-step inputs. Analyzing flight data shows that flapping motion of the main wing causes repetitive vibrations to the flyer to affect its built-in sensors. Unconstrained nonlinear optimization is applied to estimate system parameters by minimizing the difference between flight data and simulated data from the estimated linear models. As the result of system identification, the small-sized flapping aircraft used in this study shows the characteristics of the phugoid mode like fixed-aircraft. For the lateral direction, the flyer tends to turn to the right since adverse yaw happens. This paper investigates controller design and simulations based on the acquired linear models. The designed controller successfully maintains altitude and waypoint navigation enough to confirm the system identification results of the avian-type flyer system.