基于双姿态的高阶罗德里格斯参数姿态估计

Dual attitude representations based attitude estimation using high order Rodrigues parameter

针对高阶罗德里格斯参数构建姿态更新方程时存在模型奇异值以及基于四元数的姿态估计算法计算量大等问题,提出一种基于双姿态的高阶罗德里格斯参数姿态估计算法.当高阶罗德里格斯参数存在奇异时,切换到四元数姿态表示模型,避免高阶罗德里格斯参数的奇异值,之后再通过判断准则切换到高阶罗德里格参数姿态表示模型,以降低计算量.通过江试试验比较了提出算法和传统四元数无味估计器的估计精度和计算量,在计算精度基本相当的情况下,本文算法计算时间减小了26.27%.

Aiming at the problems that modeling singularity for high order Rodrigues parameter（HORP）and large amount of calculation for quaternion-based attitude estimator,a dual attitude representations based attitude estimator using HORP was proposed.In the proposed attitude estimator,the attitude representation can be switched to quaternion based model when the HORP is singular to address the HORP modeling singularty.Then,the attitude representation can be switched to HORP when the HORP is not singular through some criterions to mitigate calculation burden.The river test proves the validation of proposed attitude estimator compared with quaternion based attitude estimator.The amount of calculation is reduced by 26.27% under the same calculation accuracy.