传递对准MRP-CDKF算法  

Transfer alignment MRP-CDKF algorithm

作　　者：丁国强[1] 崔光照[1] 周卫东[2] 郝燕玲[2] 

机构地区：[1]郑州轻工业学院电气信息工程学院,河南郑州450002 [2]哈尔滨工程大学自动化学院,黑龙江哈尔滨150001

出　　处：《华中科技大学学报（自然科学版）》2011年第1期127-132,共6页Journal of Huazhong University of Science and Technology(Natural Science Edition)

基　　金：国家自然科学基金资助项目(60674087)

摘　　要：基于主、子惯导系统传递对准受控角运动和修正Rodrigues参数(MRP)的三参数描述姿态运动最小实现特点,建立了主、子惯导系统传递对准MRP非线性误差模型.根据四元数加权均值计算方法以及MRP与四元数关系,实现MRP的间接加权均值计算;根据MRP乘法定义实现MRP预测误差方差矩阵求解,进而建立修正Rodrigues参数-中心差分卡尔曼滤波(MRP-CDKF)算法.利用该算法对MRP非线性传递对准误差模型进行研究,结果表明:该算法的姿态失准角估计误差都能在10s之内收敛到0′附近;考虑杆臂向量与不考虑杆臂向量时的系统速度误差在±0.1m/s内.Based on the characteristics of the controlled angular movement for transfer alignment procedure between master and sub-intertial navigation system,and the minimal realization of modified Rodrigues parameters（MRP） representation for attitude movement,the nonlinear error model algorithm with MRP of transfer alignment was developed.According to the quaternion weighted mean calculation method and the relationship between quaternion and MRP,the method that calculates the weighted mean values of MRP indirectly.With the multiplication definition of modified rodrigues parameters,the MRP-CDKF algorithm was constructed to solve the prediction error variance matrix,and with the new algorithm studied the MRP nonlinear error model of transfer alignment system.The simulation results indicate that attitude misalignment angle estimation error of new algorithm can converge to 0′ in 10 s,and ssytem velocity error within ±0.1 m.s when considering lvever arm vector or not.

关 键 词：主惯导系统 子惯导系统 修正Rodrigues参数 传递对准 中心差分逼近原理 中心差分卡尔曼滤波算法 

分 类 号：V249.32[航空宇航科学与技术—飞行器设计] TP271[自动化与计算机技术—检测技术与自动化装置]