Date of Award

Spring 1-1-2018

Document Type

Thesis

Degree Name

Master of Science (MS)

First Advisor

J. Sean Humbert

Second Advisor

Eric W. Frew

Third Advisor

Robert B. MacCurdy

Abstract

Rejection of external disturbances is crucial for small aircraft flying in environments with dynamic flow fields, especially due to their increased sensitivity to disturbances. Fixed-wing vehicles are generally controlled using successive closure of position and velocity state feedback loops. The successive loop closure method ignores aerodynamic coupling and fails in the presence of strong gusts or aircraft damage. Existing multi-input, multi-output (MIMO) control methods may include aerodynamic coupling and provide more robust performance, but still respond slowly to high speed gusts because they feed back lower order states such as position and velocity. This study aims to develop a MIMO control methodology to feed back translational and rotational acceleration states on a simulated fixed-wing aircraft, enabling quicker rejection of disturbances. The acceleration states are linearly estimated using measurements from 12 simulated uniaxial accelerometers distributed away from the center of gravity of the vehicle. The novelty of this research is in using these acceleration states in inner feedback loops in addition to the existing autopilot to demonstrate disturbance rejection on a fixed-wing vehicle before propagation to lower order states. Using the Ttwistor small Unmanned Aircraft System (sUAS) model and robust control analysis tools, a nonlinear flight simulation with and without acceleration estimation and feedback was conducted to quantify the improvement in disturbance rejection. The resulting augmented autopilot enables the small fixed-wing UAS to fly through turbulent, gusty environments by improving disturbance rejection up to 58\% in some aircraft states. We also developed a framework for analyzing bare airframes as candidates for augmented control based on gust sensitivity and maneuverability. The results shown with the Ttwistor model demonstrate the limitations of this control approach on fixed-wing vehicles as well as the potential for even better gust rejection on smaller and more responsive platforms.

Share

COinS