飞行数据相容性检验的极大似然方法

 本文提出了一种有效的飞行仪器偏差估计的极大似然方法。为了提高对偏差估计的精度,采用了最小二乘方法对角速率,过载的测量量进行平滑,以减小系统噪声对估计精度的影响;并给出了灵敏度矩阵和初值的确定方法以保证算法的收敛速度。通过数字仿真和对我国几种不同类型飞机的飞机数据实际计算表明,提出的方法能够较准确地辨识出测量仪器的偏差,比普通极大似然法更有效。

A MAXIMUM LIKELIHOOD METHOD FOR FLIGHT TEST DATA COMPATIBILITY CHECK

An efficient maximum likelihood method for the estimation of instrumentation errors for flight is presented. In order to get high accuracy of bias estimation and to decrease the effects of measurement noises, the least-squares and shifted Chebyshev series method are applied to smoothing the measurements of roll rate, yaw rate, pitch rate, and accelerations. The methods for determining sensitivity matrix and initial values of parameters are developed and U-D factorization is devoted to computing the"ill-conditioned"matrix of parameter covariance. Simulation and actual application to two kinds of aircraft show that the method presented in this paper can give accurate estimation results of instrumentation errors of flight test system and is more efficient than the ordinary ones.