Title :
Maximum Likelihood Estimation: A method for flight dynamics - Angle of attack estimation
Author :
Lichota, Piotr ; Lasek, Maciej
Author_Institution :
Inst. of Theor. & Appl. Mech., Warsaw Univ. of Technol., Warsaw, Poland
Abstract :
In our work we are dealing with aircraft dynamic model identification. Dimensional derivatives are identified with Output Error Method - Maximum Likelihood Estimation. For finding cost function minimum Levenberg-Marquardt Algorithm is used. Fisher Information and Gradient matrices are calculated with small perturbations and central differences formulas. Initial point is estimated with Least Squares Method. Object is treated as rigid body in body fixed coordinate system. Coordinate system transformations are done with Euler rotations theorem. Equations of motion are obtained in a typical way for flight dynamics, from: Newton´s law of motion, kinematics equations and small disturbances theory. Additional terms due to process noise (turbulence) are also included.
Keywords :
gradient methods; least squares approximations; matrix algebra; maximum likelihood estimation; turbulence; vehicle dynamics; Euler rotations theorem; Fisher information; Levenberg-Marquardt algorithm; Newton law of motion; aircraft dynamic model identification; attack angle estimation; body fixed coordinate system transformation; central difference formula; cost function; dimensional derivatives; flight dynamics method; gradient matrices; kinematics equations; least squares method; maximum likelihood estimation; motion equation; output error method; process noise; small disturbances theory; small perturbations; turbulence; Aerodynamics; Aircraft; Atmospheric modeling; Equations; Mathematical model; Maximum likelihood estimation; Estimation; Flight Dynamics; Gauss; Identification; Likelihood; Probability;
Conference_Titel :
Carpathian Control Conference (ICCC), 2013 14th International
Conference_Location :
Rytro
Print_ISBN :
978-1-4673-4488-3
DOI :
10.1109/CarpathianCC.2013.6560541