期刊论文

中文

工业控制计算机

1001-182X

2019

32

2

45-47

10.3969/j.issn.1001-182X.2019.02.021

湖南省教育厅重点项目（17A182）；

本校为第一机构

针对MEMS惯性传感器噪声大、易漂移和常规姿态估计算法精度不高等问题,本着可靠和准确的原则,采用双级卡尔曼滤波DSKF（Double-State Kalman Filter）算法融合MEMS传感器数据估计姿态最优值。该算法选择四元素法表示姿态,先通过陀螺仪输出角速度估计姿态方向,然后两级阶段分别选择重力矢量和地磁矢量作为参考,利用加速度计和磁强计测量数据进行估计方向的修正。为了验证该算法准确性和实时性,将该算法与传统的卡尔曼滤波及Mahony算法进行仿真和实时估计实验比较。实验结果表明姿态估计中采用双级卡尔曼滤波算法能够输出可靠、稳定的高精度姿态数据,性能优于传统的卡尔曼滤波及Mahony算法,估计的...

Aiming at problem of high noise,drift of MEMS inertial sensors and low precision of classical attitude estimation algorithm,based on the principle of reliability and accuracy,the Double -Stage Kalman filter algorithm is used to fuse the MEMS sensor data to estimate the attitude optimal value.The algorithm selects the four-element method to represent the attitude. Firstly,the attitude direction is estimated by the gyroscope output angular velocity.Then the two-stage phase selects the gravity vector and the geomagnetic vector as the reference respectively,and uses the accelerometer and magnetome...