Comparison of on-line and off-line parameter estimation techniques using the NASA F/A-18 HARV flight data
Comparison of on-line and off-line parameter estimation techniques using the NASA F/A-18 HARV flight data
复制标题
使用 NASA F/A-18 HARV 飞行数据进行在线和离线参数估计技术的比较
DOI：
10.2514/6.2001-4261
复制
发表时间：
2001
期刊：
影响因子：
--
通讯作者：
Yongkyu Song
中科院分区：
文献类型：
--
作者：
M. Napolitano;G. Campa;B. Seanor;Yongkyu Song
近年来提出了几种在线实时应用的参数辨识（PID）技术。从历史上看，飞机参数估计是使用从专门设计的机动中记录的飞行数据进行离线的。本研究使用了NASA F/A-18高Alpha研究载具（HARV）飞机的飞行数据，并与基线风洞估计进行了比较。本文将最新开发的基于频率的PID技术与传统的极大似然方法的结果进行了比较。比较了攻角为α=20°和α=30°时机动的纵向和横向动力学。将两种估计方法的结果与现有的风洞估计结果进行了比较。符号b翼展ft c平均气动弦ft Ci气动系数1/rad或1/deg重力加速度ft/秒I惯性矩段塞/秒J成本函数-k数值系数-m飞机质量段塞p滚转速率角/秒q俯仰速率角/秒q动压力-R弧度到度的转换-q动压力磅/英尺R偏航速率角/秒S机翼面积ft时间秒t推力磅。V速度ft/秒W对角加权矩阵——已知响应的Y向量——已知输入的X矩阵——研究生研究助理，AIAA成员教授研究助理教授，AIAA成员版权所有2001作者。由美国航空航天学会授权出版。希腊α角的攻击度β侧滑角度β向量参数的估计——∆增量变化δ操纵面偏转度∇梯度,ξ参数矢量估计,θ螺旋角度φ横摇角度标准差σ估计——Σ求和ψ偏航角度频率ωrad /秒下标副翼——D阻力——dht微分水平尾巴——e电梯——L升力L滚动的时刻——lef前缘襟翼m俯仰力矩- n偏航力矩—基本机身—舵—对称副翼—尾缘襟翼—风—风轴—pv俯仰舵—Y侧向力—yv偏航舵—首字母缩写CG重心—自由度—DFT离散傅里叶变换—DFRC Dryden飞行研究中心—DTFT离散时间傅里叶变换—EE估计误差—FTR傅里叶变换回归—HARV高阿尔法研究飞行器—LS最小二乘—ML最大似然—NR牛顿-拉夫森—OBES车载励磁系统-PID参数辨识-SVD奇异值分解----
In recent years several Parameter Identification (PID) techniques have been proposed for on-line real-time applications. Historically aircraft parameter estimation has been performed off line using recorded flight data from specifically designed maneuvers. Flight data from the NASA F/A-18 High Alpha Research Vehicle (HARV) aircraft is used for this study and is compared with baseline wind tunnel estimates. This paper shows a comparison of the results of a recently developed frequency-based PID technique with the results from the traditional Maximum Likelihood method. The comparison is performed for both longitudinal and lateral-directional dynamics for maneuvers at an angle of attack of α=20° and α=30°. Results of the two estimation processes are compared with available wind tunnel estimates. Symbols b wing span ft c mean aerodynamic chord ft Ci aerodynamic coefficient 1/rad or 1/deg g gravity acceleration ft/sec I moment of inertia slug/ft J cost functional ---k numerical coefficient ---m aircraft mass slug p roll rate deg/sec q pitch rate deg/sec q dynamic pressure ---R conversion from radian to degrees ---q dynamic pressure lbs/ft r yaw rate deg/sec S wing area ft t time sec T thrust lbs. V velocity ft/sec W diagonal weighting matrix ---Y vector of known responses ---X matrix of known inputs ---Graduate Research Assistant, AIAA Member Professor Research Assistant Professor, AIAA Member Copyright 2001 by the authors. Published by the American Institue of Aeronautics and Astronautics, Inc. with permission. Greek α angle of attack deg β sideslip angle deg β vector of parameters to be estimated ---∆ incremental change ---δ control surface deflection deg ∇ gradient ---ξ parameter vector to be estimated ---θ pitch angle deg φ roll angle deg σ estimate standard deviation ---Σ summation ---ψ yaw angle deg ω frequency rad/sec Subscripts a aileron ---D drag force ---dht differential horizontal tail ---e elevator ---L lift force ---l rolling moment ---lef leading edge flap ---m pitching moment ---n yawing moment ---o basic airframe ---r rudder ---sa symmetric aileron ---tef trailing edge flap ---wind wind axis ---pv pitch vane ---Y lateral force ---yv yaw vane ---Acronyms CG Center of Gravity ---DOF Degrees Of Freedom ---DFT Discrete Fourier Transform ---DFRC Dryden Flight Research Center ---DTFT Discrete Time Fourier Transform ---EE Estimation Error ---FTR Fourier Transform Regression ---HARV High Alpha Research Vehicle ---LS Least Squares ---ML Maximum Likelihood ---NR Newton-Raphson ---OBES On-Board Excitation System ---PID Parameter Identification ---SVD Singular Value Decomposition ----