An Extended Approach to UAV Parameter Estimation using Simulated Flight Data

Abstract

Parameter estimation of aircraft aerodynamic coefficients from experimental flight data generated using dynamics of postulated equations of motion describing the aerodynamics model is presented. Since the dynamics are characterized by aerodynamic and control derivatives, then these parameters can be extracted from the flight data. The one shot least squares method based on the equation error is used to estimate the initial parameters. The response outputs of the model estimate form the basis that an update of the model parameters is sought through the use of inversing modelling, and the output error method. Model validation metrics based on statistical analysis of the parameter estimates are then used to evaluate the degree of model fit, and the predictive capability of the identified model.