级联卡尔曼滤波在初始对准中的应用

Application of Cascaded Kalman Filter in Initial Alignment

针对微惯组测量误差大,尤其是陀螺仪误差较大,不能辨别地球自转角速度,造成捷联惯性导航系统不能实现自对准的问题,采用GPS和磁强计辅助微惯组进行对准;当载体处于静止状态时,没有外力作用,加速度计仅输出重力加速度信息,可提供水平姿态角的观测信息,航向角观测由加速度计和磁强计的输出融合解算得到,通过构建级联的四个卡尔曼滤波器,分别实现水平姿态角、航向角、陀螺仪误差、失准角和加速度计零偏的估计,将估计的姿态角转换为待修正的姿态矩阵,利用估计的失准角对姿态矩阵作进一步修正,最终得到修正的姿态矩阵,完成初始对准目的。

In view of the large measurement error of the Miniature Inertial Measurement Units（MIMU）,especially the large error of the gyroscope that can＇t distinguish the self rotation speed of the earth,which cause the failure of self-alignment with only the MIMU.Therefore,the GPS and magnetometer are used to assist the MIMU in the process of alignment.When the carrier is in a static state,the accelerometer only outputs the gravity acceleration information,which can provide the observation information of the horizontal attitude angle.The direction angle observation is obtained by the output fusion solution of the accelerometer and magnetometer.And four cascaded Kalman filters are constructed to estimate the attitude angle,heading angle,gyroscope error,misalignment angle and accelerometer bias,respectively.The estimated attitude angle is converted into an attitude matrix waiting to be corrected,the attitude matrix is further modified by using the estimated misalignment angle,where completing the aim of initial alignment.