Global nonlinear parametric modelling with application to F-16 aerodynamics

Morelli, Eugene A.

doi:10.1109/acc.1998.703559

“…More recently, Morelli [21,29] and Grauer [30] applied a multi-variate polynomial model obtained from an orthogonal model structure selection to various aircraft. The latter model structure selection technique transforms the full set of candidate regressors to the orthogonal domain in order to test the significance of each parameter.…”

Section: Model Structure Selectionmentioning

confidence: 99%

Terrain Relative Navigation for Planetary Landing Using Stereo Vision Measurements Obtained from Hazard Mapping

Woicke

¹

,

Mooij

²

2017

Advances in Aerospace Guidance, Navigation and Control

View full text Add to dashboard Cite

As a result of new aviation legislation, from 2019 on all air-carrier pilots are obliged to go through flight simulator-based stall recovery training. For this reason the Control and Simulation division at Delft University of Technology has set up a task force to develop a new methodology for high-fidelity aircraft stall behavior modeling and simulation. As part of this research project, the development of a new high-fidelity Cessna II simulation model, valid throughout the normal, pre-stall flight envelope, is presented in this paper. From an extensive collection of flight test data, aerodynamic model identification was performed using the Two-Step Method. New in this approach is the use of the Unscented Kalman Filter for an improved accuracy and robustness of the state estimation step. Also, for the first time an explicit data-driven model structure selection is presented for the Citation II by making use of an orthogonal regression scheme. This procedure has indicated that most of the six non-dimensional forces and moments can be parametrized sufficiently by a linear model structure. It was shown that only the translational and lateral aerodynamic force models would benefit from the addition of higher order terms, more specifically the squared angle of attack and angle of sideslip. The newly identified aerodynamic model was implemented into an upgraded version of the existing simulation framework and will serve as a basis for the integration of a stall and post-stall model.

show abstract

“…The principles of regression analysis are well known and previously applied in many different researches in the framework of aerodynamic system identification [20,21,22]. The ordinary least squares (OLS) estimator, defined as the minimum residual…”

Section: Parameter Estimationmentioning

confidence: 99%

Identification of a Cessna Citation II Model Based on Flight Test Data

Hoek

¹

,

Visser

²

,

Pool

³

2017

Advances in Aerospace Guidance, Navigation and Control

View full text Add to dashboard Cite

As a result of new aviation legislation, from 2019 on all air-carrier pilots are obliged to go through flight simulator-based stall recovery training. For this reason the Control and Simulation division at Delft University of Technology has set up a task force to develop a new methodology for high-fidelity aircraft stall behavior modeling and simulation. As part of this research project, the development of a new high-fidelity Cessna II simulation model, valid throughout the normal, pre-stall flight envelope, is presented in this paper. From an extensive collection of flight test data, aerodynamic model identification was performed using the Two-Step Method. New in this approach is the use of the Unscented Kalman Filter for an improved accuracy and robustness of the state estimation step. Also, for the first time an explicit data-driven model structure selection is presented for the Citation II by making use of an orthogonal regression scheme. This procedure has indicated that most of the six non-dimensional forces and moments can be parametrized sufficiently by a linear model structure. It was shown that only the translational and lateral aerodynamic force models would benefit from the addition of higher order terms, more specifically the squared angle of attack and angle of sideslip. The newly identified aerodynamic model was implemented into an upgraded version of the existing simulation framework and will serve as a basis for the integration of a stall and post-stall model.

show abstract

“…The wide range of SID tools that have been developed for aircraft system identification can easily be used to analyze CFD data computed for aircraft in prescribed motion. Here we follow the global nonlinear parameter modeling technique proposed by Morelli [13] to describe the functional dependence between the motion and the computed aerodynamic response in terms of force and moment coefficients. The goal is to find a model which has adequate complexity to capture the nonlinearities while keeping the number of terms in the model low.…”

Section: System Identification Analysis (Sidpac)mentioning

confidence: 99%