The HoMM-UKF Algorithm of SINS' Attitude Estimation Model
Guoqiang DING,Duo ZHANG,Ming XIONG,Weidong ZHOU
DOI: https://doi.org/10.13705/j.issn.1671-6833.2017.01.017
2017-01-01
Abstract:In order to improve the precision requirement about the attitude control of the strap-down inertial navigation system,the high order moment matching UKF (Higher-order Moment Matching UKF,HoMMUKF) algorithm was proposed,that is to estimate the SINS' attitude parameters of based on its quaternion error model.In the recursive calculation process,for accurately approximating computational purposes,it uses high order moment matching method to calculate the average skewness value and peak value of the predicted sampling points set and their weights of the system state parameters in the view of the probability distribution.Making use of attitude quaternion method,then onlinear quaternion error model was constructed,in which model the systemnoise vector depends on system state vector,meanwhile construct its measure equation whose measurement noise vector depends on quaternion measurement vector by pseudo observation vector method was constructed,the weighted average of estimated quaternion with Lagrangian operator was calculated,the system noise variance calculation with the system noise separation algorithm was carried out,and finallyconstruct the SINS' attitude estimation HoMM-UKF algorithms simulation on SINS attitude experiment platform was designed.It can be seen that HoMM-UKF algorithm's calculation accuracy is higher than others and has better numerical stability,comparison of the UKF,and CDKF algorithms,and so the HoMM-UKF algorithm's feasibility and calculation accuracy is verified.