基于简化SSKF的SINS/GNSS紧耦合算法

摘要

针对捷联惯导(strapdown inertial navigation system,SINS)/全球卫星导航系统(global navigation satellite system,GNSS)紧耦合滤波算法中状态方程线性、观测方程非线性的特点,对超球面单型卡尔曼滤波器(spherical simplex Kalman filter,SSKF)进行了简化:采用普通卡尔曼滤波的状态矢量预测和SSKF的观测值预测及滤波更新完成滤波计算,省去了SSKF状态矢量预测时sigma点生成和对每个sigma点进行状态矢量预测和加权求和的过程,在不损失滤波精度的基础上缩短了滤波计算耗时.经过数学仿真验证,在SINS/GNSS紧耦合中,简化SSKF与SSKF几乎可到达一致的滤波估计精度,而且简化SSKF的计算耗时更短,效率更高.%The spherical simplex Kalman filter (SSKF) used in tightly coupled SINS/GNSS integration is simplified,namely simplified SSKF,due to the linear state equation but the nonlinear observation equation.Using state prediction of common Kalman filter (FK) and observation prediction of SSKF,the generating and updating of each sigma point during state prediction is omitted in simplified SSKF,which can shorten filtering time with less accuracy loss.Through mathematical simulations,simplified SSKF is proved to be more efficient than common SSKF.Besides,simplified SSKF has almost the same position (velocity) accuracy as common SSKF.