经典圆锥误差补偿算法中剩余误差估计的局限性研究  

Limitations of error estimation for classic coning compensation algorithm

作　　者：严恭敏[1] 严卫生[1] 徐德民[1] 

机构地区：[1]西北工业大学航海学院,西安710072

出　　处：《中国惯性技术学报》2008年第4期379-385,共7页Journal of Chinese Inertial Technology

基　　金：中国博士后科学基金(20070420215);水下信息处理与控制国家级重点实验室基金(9140C230206070C2306)

摘　　要：在经典圆锥误差补偿系数和剩余误差估计公式推导过程中,特别在半锥角有限小、圆锥运动频率较低和选用多子样算法时,经典的剩余误差估计方法会出现严重偏差。在不能视半锥角为无限小量情况下,文中利用推导方法提出一种新的剩余误差估计的参考公式,即圆锥误差补偿极限精度公式。利用仿真方法验证了等效旋转矢量误差的周期项对圆锥常值漂移误差有影响。研究表明,在一定的半锥角和圆锥频率的工程应用环境下,当极限精度发挥作用时,提高圆锥误差补偿精度的唯一途径是缩短补偿周期,而不像经典的结论那样——通过提高子样数几乎能无限提高圆锥误差补偿精度。In the deducing of the classic coning compensation algorithm coefficients and their residual error estimation formulas, the classic error estimation formula could cause serious errors in the case of insufficient small coning half-angle amplitude, low coning frequency and multi-samples algorithm. So a new error estimation formula, i.e. accuracy limitation formula of coning compensation was deduced even if the coning angle is not infinitesimal. Moreover, simulation tests verify that the periodic components of equivalent rotation vector affect coning compensation errors. The studies reveal that if the accuracy limitation formula acts on coning compensation algorithm under a large attitude updating period, the approach of improving coning compensation accuracy is to shorten the equivalent rotation vector calculation period, while the classic means is just by increasing the number of samples.

关 键 词：圆锥误差补偿 等效旋转矢量 四元数 角增量 捷联惯导系统 

分 类 号：U666.1[交通运输工程—船舶及航道工程]