Frequency Domain System Identification of a Small Flying-Wing UAS

Abstract

This paper focuses on system identification of a small, flying-wing UAS using the frequency response method. A flight test procedure is designed to address the unique challenges encountered when conducting system identification for a small flying-wing UAS with elevon controls. These challenges include increased susceptibility to atmospheric disturbances, limited yaw maneuverability, and visual line-of-sight safety requirements. Frequency sweeps are used as control inputs to excite the longitudinal and lateral-directional dynamics over a designed frequency range. Reduced-order transfer functions are first identified to gain initial information on key dynamics and to provide comparison with different models. Then, decoupled longitudinal and lateral-directional state space models are identified from flight data. The models are validated in the time-domain through comparison with doublet maneuver flight data, showing an excellent fit between the dynamic models and flight data. Finally, nondimensional stability and control derivatives and their confidence intervals are computed from the state space models for comparison with other modeling methods.