基于MARG传感器的自主姿态测量新算法  

A New Algorithm of Autonomous Attitude Determination Using MARG Sensor

作　　者：傅建国[1] 王孝通[1] 金良安[1] 杨常青[1] 

机构地区：[1]海军大连舰艇学院航海系,大连116018

出　　处：《中国惯性技术学报》2004年第4期51-56,共6页Journal of Chinese Inertial Technology

基　　金：军队武器装备科研项目(H1022003166)

摘　　要：针对基于三轴陀螺、三轴加速度计以及三轴磁强计组成的传感器组合(MARG传感器)的姿态测量算法存在的不足和限制,以自主姿态测量为前提,提出了一种新的基于四元数的优化递推估计算法。这一算法能够在线估计补偿陀螺漂移误差,不再需要对系统方程进行线性化处理。与基于EKF的方法相比,在同等计算复杂度下,该算法具有更高的精度和更好的稳定性。算法本身内蕴了对四元数单位长度的限制,不再需要任何归一化处理。仿真结果显示出了该算法具有令人满意的效果,利用水上漂浮试验初步验证了该算法的稳定性。A new quaternion-based autonomous attitude estimation algorithm using multiple Magnetic Angular Rate Gravity (MARG) sensor units is designed to address the shortage of primary algorithms. The presented algorithm can estimate and compensate gyro drift on-line. The new design avoids the linearization of the nonlinear system equations, and has higher theoretical accuracy and stability than the EKF-based or EKF-like algorithms. The unit length limitation of the quaternion is guarantied inherently in the new algorithm. The simulation results show that the new design is reasonable and its performance is satisfying. The convergence and stability of the algorithm are validated elementarily by floating test on the water.

关 键 词：船舶导航 自主姿态测量 非线性递推估计 Unscented变换 Unscented滤波 四元数 

分 类 号：U666.12[交通运输工程—船舶及航道工程]