基于修正的卡尔曼滤波的姿态估计算法研究

研究惯导系统的稳定性问题,其中微惯性测量单元(MIMU)可以为捷联惯导系统提供实时的姿态和航向信息。研究姿态估计提高导航精度,由于陀螺漂移引起姿态误差,单独使用MIMU使姿态精度差。为了克服陀螺误差随时间积累不断增大,无法长时间提供稳定的姿态的缺点,提出采用磁强计修正的卡尔曼滤波四元数姿态估计算法。算法以姿态四元数为状态向量,通过四元数更新方程建立离散滤波状态方程,将加速度计和磁强计输出的六维数据转化为四元数的量测值建立量测方程,有效减少了计算量,补偿陀螺的漂移误差带来的影响。仿真结果表明改进算法提高了捷联惯导系统的精度和稳定性。

Research on Attitude Estimation Algorithm Based on Kalman Filer

The real-time attitude and heading of Strap-down Inertial Navigation System(SINS) can be supplied by Micro Inertial Measurement Unit(MIMU),so the research on attitude estimate algorithm is the important means to improve accurancy.MIMU cannot provide a steady attitude alone for a long time as the error is increasing along with time accumulation.A quaternion based on modified Kalman filter attitude estimation algorithm for magnetometer was presented to improve the precision.Firstly,attitude quaternion was selected as state vector,then the state equation was established through quaternion updating equation.The quaternion measurement value was transformed by the 6th-dismention data of accelermeter and magnetometer,which can efficiently simplify computation and compensate the drift of gyroscope.Simulation illustrates that the algorithm improves accuracy and stability of SINS.