中文摘要：

惯性—地磁组合由于不需要外部参考源而可获得运动体姿态,因而在大空间范围内的人体运动跟踪领域获得广泛应用.目前适用于惯性—地磁组合的姿态测量算法很多,然而在传感器随机误差及运动体线加速度干扰的影响下,这些算法的静动态性能各异.详细分析了五种姿态解算算法,包括线性卡尔曼算法、状态方程非线性的卡尔曼算法、观测方程非线性的卡尔曼算法、REQUEST算法以及Optimal-REQUEST算法,结果表明,观测方程非线性的卡尔曼算法以及Optimal-REQUEST算法兼具良好的静动态性能,然而前者受初始估计影响严重,后者存在漂移现象,因而一个很好的建议是将两者进行整合,以构成一种新的具有更好静动态性能的姿态解算算法.

英文摘要：

Nowadays,inertial-magnetic units（IMU） have been widely used in human body movement tracking,for they do not rely on any external reference for attitude determination.There are many attitude algorithms for IMU,however,they have different static and dynamic performances under the influence of the random error of the sensors and the interference of the linear acceleration of the moving objects.this paper deeply analyzed five algorithms,which included the linear Kalman filter （LKF）,the extended Kalman filter（EKF） with its state equation being nonlinear（EKF1）,the EKF with its observation equation being nonlinear（EKF2）,the REQUEST,and the optimal-REQUEST,and the results show that,the EKF2 and the optimal-REQUEST are the best,regarding their static and dynamic performances,however,the outputs of the former are affected heavily by the initial estimation,and the outputs of the latter drift with time,so a good idea is to integrate two of them with the aim to form a new algorithm which does not have any defects mentioned above.