Robust Aerodynamic Parameter Estimation of Unmanned Aircraft Based on Two-step Identification

Abstract

This paper presents the estimation of stability and control derivatives of an unmanned aircraft. The aerodynamics are described using regressors composed of velocity, angular rates, flow angles and control surface deflections. The flight data is generated from numerical simulation of postulated equations of motion describing the aerodynamics model. Least squares based on the equation error method is used to estimate the parameters representing the different force and moment aerodynamic coefficients. Statistical analysis is done on the estimates to determine the accuracy and adequacy of the estimates to describe the aerodynamic model. A dynamic simulation based on the identified aerodynamic model is used to improve the parameter estimates through regression of the errors between the flight data and the model response. The aircraft under consideration is a scaled Yak-54 fixed wing unmanned aerial vehicle.