捷联惯导系统快速初始对准的多级组合滤波实现

Fast SINS Initial Alignment by Integrated Filtering Methods

捷联惯导系统初始对准中利用卡尔曼滤波器进行状态估计时,其数学平台的东向和北向水平失准角估计速度很快,而天向方位失准角的收敛很慢。为了解决这个问题,在基本卡尔曼滤波的基础上引入了多级组合滤波的思想,利用对捷联矩阵的修正和对天向方位失准角φU估计的改进产生了新的信息,形成了两级卡尔曼滤波,并且与小波软阈值法相结合,实现了捷联惯导系统的快速初始对准。

In SINS initial alignment, east and north level error angles can be perfectly acquired by Kalman filtering, whereas the estimation effect of azimuth error angle φU was bad. The multi-stage integrated filtering was proposed. The new information of attitude matrix and the new estimation method of azimuth error angle φU was used by the new filtering, also the de-nosing by soft thresholding methods was utilized to eliminate the noise. The simulation shows that the time of initial alignment reduces greatly, the new fast estimation method performs better.