Fuzzy-Logic Modeling of a Rolling Unmanned Vehicle in Antarctica Wind Shear

Abstract

of wind shear are investigated using the recorded test data during a flight from the Pegasus white ice runway in Antarctica. The main purpose of this paper is to determine the reasons and characteristics of roll oscillations for improvement of autopilot design and ground operator training. The Meridian flight test is conducted with the necessaryrangeconstraint, which precludes steadylevel flight.The measured orestimated data,such as theangle of attack and sideslip angle, are filtered through a nonlinear compatibility analysis. The estimated aerodynamic coefficients are modeled as implicit functions of filtered flight variables in a concept of aerodynamic model identification, instead of parameter identification. A model identification method called fuzzy-logic modeling is used to set up the aerodynamic models without assuming the functional relations. The aircraft’s oscillatory stability and control derivatives are found to be significantly impacted by the environmental disturbances. The exhibited highly oscillatory motion in the manual-control mode is analyzed and identified as the pilot-induced oscillation with a new energy method. It is shown that the pilot-induced oscillation is identified as the motion when the motion energy is increasedbythecontrolinput.Intheautopilot flight,themotionisdeterminedtobewingrock,whichisexcitedbythe pilot-induced oscillation in the manual mode and sustained by the interaction between the wing wake and the V tail.