SINS大失准角初始对准的改进算法研究

SINS Initial Alignment of Big Deviation Angle Improved Algorithm Research

在捷联导航系统优化控制的研究中,捷联惯导系统(SINS)初始对准的误差方程存在是非线性,对于静基座初始对准造成误差较大。传统采用的方法是将失准角视为小角度,可将误差模型线性化,利用KF完成静基座的初始对准。但是对于动基座大失准角来说,多采用非线性滤波方式来解决,建立误差模型,并采用UKF滤波进行数据融合。由于非线性模型的噪声参数未知等原因,常规的UKF可能会出现滤波发散现象,为解决上述问题,提出采用自适应方法的UKF来对建立的非线性模型进行滤波。仿真结果表明,在大失准角情况下,采用非线性模型的AUKF比UKF滤波具有更好的对准精度和更快的收敛速度,可为优化设计提供参考。

In the research of optimal control of the strapdown inertial navigation system （SINS） , because of the system＇s initial alignment error equation is nonlinear, there are great errors to the base alignment for initial alignment. Traditionally, we treat deviation angle as small angle and use Kalman filter（KF） to accomplish initial alignment. But, in the case of move base big deviation angle, we usually use nonlinear filtering method to solve the problem. Firstly, we built the error models. Then we used Unscented Kalman Filter（UKF）. Since the noise parameter of the nonlinear model, normal UKF maybe invalid. To solve the above problems, this paper proposed a new UKF method based on self-adaption. The simulation results show that in the case of big deviation angle, the self-adaption UKF （AUKF） has better accuracy and faster convergence rate than UKF.