一种可观测度分析方法及在传递对准中的应用

Observable degree analysis method and its application in transfer alignment

针对卡尔曼滤波器中状态可观测度的定量分析问题,提出了一种新的可观测度分析方法。利用无系统噪声输入的动态系统初始状态的加权最小二乘估计,推导了动态系统初始状态估计误差的传递方程,将状态变量的可观测度指标定义为估计误差传递矩阵的对角线元素,从系统初始状态估计误差衰减角度定义了可观测度。推导了考虑子惯导安装矩阵的姿态观测方程,将机翼变形视为观测噪声,构建了速度加姿态匹配的传递对准模型。将可观测度分析方法应用到该传递对准模型中,结果表明:水平加速计零偏可观测度低,估计误差大,而其他状态可观测度高,估计误差小。

Focusing on the problem of observable degree quantitative analysis for Kalman filter, a new method for observable degree analysis is presented. The system initial state estimation error transfer equation is derived for weighted least squares estimation of dynamic system with no system noise input. The diagonal element of estimation error transfer matrix is defined as the system state observable degree index. The observable degree is defined from the estimation error attenuation viewpoint. The attitude measurement equation is derived considering the slave inertial navigation system＇s installation matrix. The transfer alignment of velocity plus attitude match model is built, taking the wing flexure as the measurement noise. The new observable degree analysis method is applied to this transfer alignment model. The simulation results show that level accelerometer bias observable degree is low and estimation error is large, while other states observable degree is high and estimation error is small.