Progress in astronautics and aeronautics. V. 216, Flight vehicle system identification : a time domain methodology

This valuable volume offers a systematic approach to flight vehicle system identification and exhaustively covers the time domain methodology. It addresses in detail the theoretical and practical aspects of various parameter estimation methods, including those in the stochastic framework, and focuses on nonlinear models, cost functions, optimization methods, and residual analysis. A pragmatic and balanced account of pros and cons in each case is provided. The book also presents data gathering and model validation and covers both large-scale systems and high-fidelity modeling. Real-world problems dealing with a variety of flight vehicle applications are addressed and solutions are provided. Examples encompass such problems as estimation of aerodynamics, stability, and control derivatives from flight data, flight path reconstruction, nonlinearities in control surface effectiveness, stall hysteresis, unstable aircraft, and other critical considerations. Beginners, as well as practicing researchers, engineers, and working professionals who wish to refresh or broaden their knowledge of flight vehicle system identification, will find this book highly beneficial. Based on years of experience, the author also provides recommendations for overcoming problems likely to be faced in developing complex nonlinear and high-fidelity models, and the book can help the novice negotiate the challenges of developing highly accurate mathematical models and aerodynamic databases from experimental flight data. - Data and information appearing in this book are for informational purposes only. AIAA and the author are not responsible for any injury or damage resulting from use or reliance, nor do AIAA and the author warrant that use or reliance will be free from privately owned rights. Front Matter 1 Table of Contents 3 1. Introduction 8 1.1 What is System Identification? 9 1.2 Model Characterization 12 1.3 Interdisciplinary Flight Vehicle Modeling 13 1.4 Why System Identification? 15 1.5 Parameter Estimation in Flight Mechanics 16 1.6 Estimation Techniques of the Past 18 1.7 Modern Methods of Aircraft Parameter Estimation 19 1.8 General Aspects 21 1.8.1 The Aim 21 1.8.2 The Literature 21 1.8.3 The Layout 22 1.8.4 Software and Flight Data 26 References 27 2. Data Gathering 31 2.1 Introduction 31 2.2 Flight Testing and Maneuvers 32 2.2.1 Flight Testing for Performance Evaluation 33 2.2.2 Flight Testing for System Identification 35 2.3 Optimal Input Design 39 2.3.1 Input Design by Estimation Error Analysis 39 2.3.2 Design of Multistep Input Signals 42 2.4 Scope of Flight Testing 52 2.5 Flight Test Instrumentation and Measurements 55 2.5.1 Digital Filter 59 2.5.2 Numerical Differentiation 59 2.6 Concluding Remarks 60 References 62 3. Model Postulates and Simulation 65 3.1 Introduction 65 3.2 Model Description 66 3.3 Extensions of the Mathematical Models 67 3.4 Retarded Systems 69 3.5 Linearized Models 73 3.5.1 Identifiability of Aerodynamic Derivatives and Constant Parameters 73 3.5.2 Models with Lumped Bias Parameters 74 3.5.3 Numerical Approximation of System Matrices 75 3.6 Pseudo-Control Inputs 75 3.7 Treatment of Initial Conditions 77 3.8 Simulation 77 3.8.1 Numerical Integration Methods 78 3.8.2 Integration of Linear Systems 81 3.9 Concluding Remarks 82 References 83 4. Output Error Method 85 4.1 Introduction 85 4.2 The Principle of Maximum Likelihood Estimation 86 4.3 Properties of Maximum Likelihood Estimates 89 4.4 The Maximum Likelihood Function for Estimation of Parameters in Dynamic Systems 90 4.5 Basics of Cost Function Optimization 92 4.5.1 Known Measurement Noise Covariance Matrix 92 4.5.2 Unknown Measurement Noise Covariance Matrix 92 4.6 Gauss-Newton Algorithm 94 4.7 Method of Quasi-Linearization 96 4.8 System Response and Sensitivity Coefficients 97 4.9 Automatic Gradient Computation 100 4.10 Step Size Control 101 4.10.1 Heuristic Approach 101 4.10.2 Line Search 102 4.10.3 Dominant Directions 102 4.11 Bounded-Variable Gauss-Newton Method 104 4.12 Constrained Gauss-Newton Method Using the Interior-Point Algorithm 106 4.13 Levenberg-Marquardt Method 109 5. Filter Error Method 111 5.1 Introduction 111 5.2 Filter Error Method for Linear Systems 113 5.3 Process Noise Formulations 115 5.3.1 Natural Formulation 115 5.3.2 Innovation Formulation 116 5.3.3 Combined Natural cum Innovation Formulation 118 5.4 Filter Error Algorithm 118 5.4.1 Solution to Riccati Equation 119 5.4.2 Parameter Update 120 5.5 Filter Error Method for Nonlinear Systems 125 5.5.1 Steady-State Filter 126 5.5.2 Time-Varying Filter 131 6. Equation Error Methods 134 6.1 Introduction 134 6.2 Least Squares Method 135 6.2.1 Properties of Least Squares Estimates 139 6.2.2 Practical Considerations 142 6.2.3 Applicability to State Space Models 144 6.3 Weighted Least Squares Method 145 6.4 Nonlinear and Multi-Output Regression 146 6.5 Total Least Squares 148 6.6 Instrumental Variable Method 151 6.7 Data Partitioning 153 6.8 Model Structure Determination 154 6.8.1 Stepwise Regression 156 6.8.2 Statistical Test Criteria 158 6.8.3 Practical Aspects 160 6.9 Examples 161 6.9.1 Flight-Derived Aerodynamic Force and Moment Coefficients 161 6.9.2 Estimation of Aerodynamic Parameters of a Linear Model 164 6.9.3 Estimation of Aerodynamic Parameters from a Nonlinear Model 172 6.10 Concluding Remarks 173 References 174 7. Recursive Parameter Estimation 176 7.1 Introduction 176 7.2 Least Squares-Based Recursive Methods 179 7.2.1 Recursive Least Squares 179 7.2.2 Recursive Weighted Least Squares 182 7.2.3 Locally Weighted Regression Method 183 7.2.4 Fourier Transform Regression 186 7.3 Filtering Methods 191 7.3.1 Extended Kalman Filter 192 7.3.2 Unscented Kalman Filter 194 8. Artificial Neural Networks 200 8.1 Introduction 200 8.2 Basics of Neural Network Processing 203 8.3 Training Algorithms 205 8.3.1 Forward Propagation 206 8.3.2 Standard Back-Propagation Algorithm 207 8.3.3 Back-Propagation Algorithm with Momentum Term 209 8.3.4 Modified Back-Propagation Algorithm 210 8.4 Optimal Tuning Parameters 211 8.5 Extraction of Stability and Control Derivatives from Trained FFNN 213 8.6 FFNN Software 214 8.7 Examples 216 8.7.1 Modeling of Lateral-Directional Aerodynamics 216 8.7.2 Modeling of Longitudinal Data with Process Noise 218 8.7.3 Quasi-Steady Stall Modeling 220 8.8 Concluding Remarks 224 References 225 9. Unstable Aircraft Identification 229 9.1 Introduction 229 9.2 Basics of Unstable Aircraft Identification 231 9.3 Least Squares Method 233 9.3.1 Detection of Data Collinearity 234 9.3.2 Estimation in the Presence of Data Collinearity 235 9.4 Total Least Squares Method 237 9.5 Combined Output Error and Least Squares Approach 238 9.6 Equation Decoupling Method 238 9.7 Eigenvalue Transformation Method 240 9.8 Filter Error Method 242 9.9 Extended and Unscented Kalman Filters 243 9.10 Output Error Method 244 9.11 Output Error Method with Artificial Stabilization 244 9.12 Multiple Shooting Method 245 9.13 Output Error Method in Frequency Domain 247 9.14 Separate Surface Excitation 248 9.15 Programming Considerations 250 9.16 Examples 251 9.16.1 Simulated Unstable Aircraft Response Data 251 9.16.2 Identification of X-31A Lateral-Directional Motion 259 9.17 Concluding Remarks 265 References 266 10. Data Compatibility Check 268 10.1 Introduction 268 10.2 Kinematic Equations 269 10.3 Flight Path Reconstruction Techniques 277 10.3.1 Deterministic Approach 277 10.3.2 Stochastic Approach 278 10.4 Estimation-before-Modeling Approach 283 10.5 Example 287 10.6 Calibration of Five-Hole Flow Angle Probe 291 10.7 Calibration of Static Pressure Ports 296 10.7.1 Tower Flyby and Measurements 297 10.7.2 Data Analysis 298 10.8 Wind-Box Maneuver Technique 301 10.9 Concluding Remarks 304 References 305 11. Model Validation 308 11.1 Introduction 308 11.2 Statistical Accuracy of Parameter Estimates 309 11.3 Residual Analysis 311 11.3.1 Goodness of Fit 311 11.3.2 Theil's Inequality Coefficient and Decomposition of Fit Error 312 11.3.3 Test for Whiteness 314 11.4 Inverse Simulation 315 11.5 Model Plausibility 316 11.6 Model Predictive Capability 319 11.7 Range of Model Applicability in Frequency Domain 322 11.8 Concluding Remarks 325 References 325 12. Selected Advanced Examples 328 12.1 Introduction 328 12.2 Modeling of Transit Time Lag Effects 329 12.2.1 Separation of Pitch Damping Derivatives 329 12.2.2 Aerodynamic Effects of Direct Lift Control Flaps 336 12.2.3 Modeling of Speed Brake Effect 339 12.3 Aerodynamic Effects of Landing Gear 342 12.4 Control Surface Malfunction Effects 344 12.5 Unsteady Aerodynamics Modeling 347 12.6 Quasi-Steady Stall Modeling 351 12.6.1 ATTAS Stall Modeling 353 12.6.2 Dornier 328 Stall Modeling 355 Epilogue 358 References 360 Appendices 361 Appendix A: Power Spectrum of a Multistep Input Signal 361 References 363 Index 364 A 364 B 367 C 369 D 371 E 373 F 375 G 380 H 381 I 381 J 383 K 383 L 384 M 387 N 391 O 393 P 394 Q 398 R 398 S 401 T 406 U 408 V 409 W 409 X 410 Z 410

This valuable volume offers a systematic approach to flight vehicle system identification and exhaustively covers the time domain methodology. It addresses in detail the theoretical and practical aspects of various parameter estimation methods, including those in the stochastic framework and focusing on nonlinear models, cost functions, optimization methods, and residual analysis. A pragmatic and balanced account of pros and cons in each case is provided. The book also presents data gathering and model validation, and covers both large-scale systems and high-fidelity modeling. Real world problems dealing with a variety of flight vehicle applications are addressed and solutions are provided. Examples encompass such problems as estimation of aerodynamics, stability, and control derivatives from flight data, flight path reconstruction, nonlinearities in control surface effectiveness, stall hysteresis, unstable aircraft, and other critical considerations.

Beginners, as well as practicing researchers, engineers, and working professionals who wish to refresh or broaden their knowledge of flight vehicle system identification will find this book highly beneficial.

Based on years of experience, the author also provides recommendations for overcoming problems likely to be faced in developing complex nonlinear and high-fidelity models, and the book can help the novice negotiate the challenges of developing highly accurate mathematical models and aerodynamic databases from experimental flight data. Software that runs under MATLAB® and sample flight data are provided to assist the reader in reworking the examples presented in the text. The software can also be adapted to the reader's own interests.

MATLAB(R) is a registered trademark of The MathWorks, Inc.

This valuable volume offers a systematic approach to flight vehicle system identification and covers exhaustively the time-domain methodology. It addresses in detail the theoretical and practical aspects of various parameter estimation methods, including those in the stochastic framework and focusing on nonlinear models, cost functions, optimization methods, and residual analysis. A pragmatic and balanced account of pros and cons in each case are provided. It also presents data gathering, model validation and covers both large scale systems and high fidelity modeling. Real world problems dealing with a variety of flight vehicle applications are addressed and solutions are provided. Examples encompass such problems as estimation of aerodynamics, stability, and control derivatives from flight data, flight path reconstruction, nonlinearities in control surface effectiveness, stall hysteresis, unstable aircraft, and other critical considerations. Beginners, as well as practicing researchers, engineers, and working professionals who wish to refresh or broaden their knowledge of flight vehicle system identification will find this book highly beneficial. Based on years of experience, the book also provides recommendations for overcoming problems likely to be faced in developing complex nonlinear and high fidelity models and can help the novice negotiate the challenges of developing highly accurate mathematical models and aerodynamic databases from experimental flight data. Software that runs under MATLAB[registered] and sample flight data are provided to assist the reader in reworking the examples presented in the text. The software can also be adapted to the reader's own interests. Content: Front Matter • Interactive Graphs Table (143) • Preface • Table of Contents 1. Introduction 2. Data Gathering 3. Model Postulates and Simulation 4. Output Error Method 5. Filter Error Method 6. Equation Error Methods 7. Recursive Parameter Estimation 8. Artificial Neural Networks 9. Unstable Aircraft Identification 10. Data Compatibility Check 11. Model Validation 12. Selected Advanced Examples Epilogue Appendices Index Offers a systematic approach to flight vehicle system identification and covers the time-domain methodology. This book addresses the theoretical and practical aspects of various parameter estimation methods, including those in the stochastic framework and focuses on nonlinear models, cost functions, optimization methods, and residual analysis.