基于MIMU和磁力计的姿态更新算法研究

针对传统人体姿态解算算法中存在MEMS陀螺误差发散快的问题，提出一种基于微惯性测量单元( MIMU)及磁力计信息融合的姿态解算算法。该算法利用互补滤波结合PI调节控制完成陀螺零偏校正，然后在加速度计和磁强计的辅助校正下，通过EKF( Expand Kalman Filter)滤波器更新四元数法实现陀螺姿态解算。本算法采用MPU9150传感器模块完成测试实验，实验中对比分析了单独扩展卡尔曼滤波算法与本算法的滤波效果。实验结果表明，本算法能够有效地抑制陀螺的发散，实现稳定地输出高精度姿态数据。

MIMU update algorithm based on the posture and magnetometer

Since the conventional human attitude algorithms suffer from the fast divergence of MEMS gyro errors,this paper proposes a novel attitude algorithm based on the information fusion of micro inertial measurement unit ( MIMU) and magnetometers. We first rely on the filter combined with PI regulation to conduct gyro zero slant cor-rection,and then in the condition of accelerometer and magnetometer auxiliary correction,realize the gyro attitude algorithm by using the extended Kalman filter( EKF) to update quaternion. By selecting the MPU9150 sensor mod-ule,the extensive experiments are provided to compare the performance of the conventional single EKF and the pro-posed algorithm in filtering. The results demonstrate that the proposed algorithm can significantly restrain the diver-gence of MEMS gyro errors,and consequently output the stable and highly-accurate attitude data.