摘要
构建了基于MEMS陀螺仪、加速度计及磁强计的姿态测量系统。研究了基于四元数的扩展卡尔曼滤波算法。通过四元数微分方程和陀螺噪声误差建立了卡尔曼状态方程,利用加速度计和磁强计的输出数据,采用梯度下降法计算出测量四元数。用从实际的惯性测量单元采集的数据对滤波器进行测试,结果证明航向角在通过卡尔曼滤波器后,有效的抑制了姿态角的发散,使得航向角姿态角误差小于3°,该算法能够实现小型无人机的高精度姿态解算。
Attitude measurement system is constructed based on MEMS gyroscope, accelerometer and mag- netometer, and the extended Kalman filtering algorithm is studied based on quaternion. Kalman state equation is established through the quaternion differential equation and the gyro noise error. Using the output data of accelerometer and magnetometer, measurement quaternion is worked out with the gradient descent method. The filter is tested with data collected from actual inertial measurement unit, and the results prove that head- ing angle through Kalman filter can effectively inhibit the divergence of attitude angle, with the heading an- gle error less than 3°. The algorithm can achieve high precision attitude algorithm of small drone.
出处
《沈阳航空航天大学学报》
2013年第6期53-57,共5页
Journal of Shenyang Aerospace University
基金
辽宁省重点实验室项目资助(项目编号:LS2010117)
关键词
扩展卡尔曼滤波
姿态解算
梯度下降法
extended kalman filter
attitude algorithm
gradient descent method
