New algorithm of maximum likelihood parameter estimation for flight vehicles

This paper proposes a new algorithm for the maximum likelihood parameter estimation problems with both process and measurement noise. Maximum likelihood estimators commonly minimize a cost function indicating the difference between measured and estimated responses. In the new formulation, equality constraints are added to the minimization to eliminate the convergence problems of previous formulations. In addition, a Sequential Quadratic Programming (SQP) method with a Gauss-Newton approximation is adopted to efficiently solve the constrained minimization problem. This paper also presents the applications of the proposed algorithm. The algorithm is first applied to a simple scalar system, and the stability and control derivatives are estimated from simulated and real flight data. In the analyses, it is verified that the algorithm remarkably improves convergence and provides reasonable estimates. Nomenclature Abbreviations and Acronyms FADS Flush Air Data System HYFLEX Hypersonic Flight Experiment IMU Inertial Measurement Unit MMLE Modified Maximum Likelihood Estimator NAL National Aerospace Laboratory, Japan NASA National Aeronautics and Space Administration NASDA National Space Development Agency of Japan SQP Sequential Quadratic Programming Symbols A, B continuous system matrices c equality constraint vector CR Cramer-Rao bound d; search direction vector d2 correction vector F penalty function g gravitational acceleration, m/s Associate Senior Engineer, Office of Space Transportation Systems, formerly with the National Aerospace Laboratory, Japan. Member AIAA. Copyright © 1997 by the American Institute of Aeronautics and Astronautics, Inc. All rights reserved. H I J K L', m N n P x p p, Q q RI r R2 S t u V v w X Y, z a P A 8 8 , M, N'

[1]  Kenneth W. Iliff,et al.  Extraction of stability and control derivatives from orbiter flight data , 1993 .

[2]  R. Maine,et al.  Formulation and implementation of a practical algorithm for parameter estimation with process and measurement noise , 1980 .

[3]  M. Sahba.  Globally convergent algorithm for nonlinearly constrained optimization problems , 1987 .

[4]  Eugene A. Morelli,et al.  Accuracy of Aerodynamic Model Parameters / Estimated from Flight Test Data , 1997 .

[5]  R. E. Maine,et al.  Application of parameter estimation to aircraft stability and control: The output-error approach , 1986 .

[6]  Yonathan Bard,et al.  Comparison of Gradient Methods for the Solution of Nonlinear Parameter Estimation Problems , 1970 .

[7]  K. W. Iliff,et al.  Parameter Estimation for Flight Vehicles , 1989 .

[8]  Masao Fukushima,et al.  A successive quadratic programming algorithm with global and superlinear convergence properties , 1986, Math. Program..

[9]  D. Mayne,et al.  A surperlinearly convergent algorithm for constrained optimization problems , 1982 .

[10]  C. Loan,et al.  Nineteen Dubious Ways to Compute the Exponential of a Matrix , 1978 .

[11]  R. E. Maine,et al.  A FORTRAN program for determining aircraft stability and control derivatives from flight data , 1975 .

[12]  R. V. Jategaonkar,et al.  Algorithms for aircraft parameter estimation accounting for process and measurement noise , 1989 .

[13]  A. Laub,et al.  On the numerical solution of the discrete-time algebraic Riccati equation , 1980 .

[14]  R. Mehra,et al.  Computational aspects of maximum likelihood estimation and reduction in sensitivity function calculations , 1974 .

[15]  Raman K. Mehra,et al.  Maximum likelihood identification of aircraft stability and control derivatives , 1974 .