Research On CQKF In SINS Initial Alignment For Large Misalignment Angles

Kalman Filter is the most classical approach to be applied to linear systems,but in the real system,the vast majority of cases are nonlinear,so Kalman Filter and the traditional linear model of the filter can not estimate the system very completely.Therefore the establishment of nonlinear model and the nonlinear filtering,is more suitable the requirements of modern system on the performance of filter.While the nonlinear filtering method which approximate the actual distribution of sampling strategy usingdeterministic system has been widely attented and applied,because it do not need the process of linearization for nonlinear systems.Cubature Quadrature Kalman Filter(CQKF)algorithms are relatively new this year,its features are high precision,more adapted to the nonlinear systems.CQKF algorithm is getting more attention and become effective filtering algorithm for dealing with nonlinear filtering problem.Strapdown inertial navigation System technology has a very important role in aviation,navigation,guidanceand other fields.Initial alignment is a key technique in strapdown inertial navigation,the aligning accuracy and time directly influences the performance of the system.Traditional filtering methods only take into account the azimuth angle is a small angle and linearized the system then completes the initial alignment.However,the azimuth misalignment angle is always the large angle.Therefore the establishment of large azimuth misalignment angle model is an important way to improve the performance of the system.Firstly,this paper ntroduces the principle and characteristics of the traditional Kalman filter,but the algorithm can only be applied to linear model,therefore,this paper focuses on the analysis of the classic Cubature Kalman Filter(CKF)algorithm.CKF has some advantages including less possibility for divergence,higher precision.Secondly,This paper focuses on the study of basic theory and the derivation process of CQKF,and summarizes the characteristics of the algorithm.CQKF is the evolution of CKF,in the fully inherited the advantages at the same time,and further improves the filtering precision.CKF in the radial integral part has made only one order of accuracy,in order to solve this problem,CQKF using the Gauss-Laguerre quadrature code will be integral precision of this part of the increased to two order.Thirdly,noise have a negligible effect on the performance of these filtering methods.excessive dependence on the accuracy of prior noise statistics for all of the above filtering algorithm is an inevitable question.In practice,most of the noise statistics system is inaccurate and even unknowable.Aiming at this problem,this paper researches the principle of adaptive Sage-Husa filter,combined with the CQKF.the Adaptive CQKF algorithm is proposed to solve the problem of noise.Finally,the traditional initial alignment will linearize model,this method has problems of instability and low accuracy.Aiming at the problems this paper introduce and use a large azimuth misalignment angle initial alignment model,and the CQKF algorithm is applied to the system,to obtain higher alignment accuracy.This paper will apply noise adaptive CQKF algorithm to the strapdown inertial navigation large azimuth misalignment angle of initial alignment system,obtain the higher alignment accuracy,and the problem of the noise also obtained the very good solution.The simulation results prove the filter algorithm is effective.