Research On Initial Alignment Method Of Rotating Carrier Integrated Navigation System

While rotating system improves flight stability of the carrier,it brings great challenges to achieve high precision alignment and navigation.The new generation of rotating carriers mostly use(Global Navigation Satellite System,GNSS)/(Inertial Navigation System,INS)integrated navigation system to obtain flight parameters such as attitude,speed and position.The premise of its subsystem,INS,is to complete the initial alignment that means to get the initial relationship of body frame relative to the reference navigation frame.Therefore,there is an urgent need to find a method to achieve fast and accurate initial alignment of rotating carriers and provide a basis for navigation.This paper studies the initial alignment algorithm of the rotating carrier integrated navigation system,including the in-flight coarse alignment algorithm and the in-flight fine alignment algorithm.The main content of this paper is as follows:Firstly,the basic principle of strapdown inertial navigation and the kinematic characteristics of the rotating carrier are described.According to the kinematic characteristics of the rotating carrier,the ballistic model of the rigid body is designed to obtain the ballistic simulation data,which provides the data basis for the subsequent initial alignment algorithm.Then,aiming at the problem of fast coarse alignment,multi-vector in-flight alignment method based on dynamic model constraints is proposed.The principle of coarse alignment,vector construction and vector attitude determination are described.After that,the effectiveness of the multi-vector in-flight alignment method in the ballistic environment is analyzed.On the basis of the traditional multi-vector in-flight alignment algorithm,a ballistic kinematic estimation algorithm based on EKF is introduced.Results indicated that,the method proposed in this paper can complete the coarse alignment in 7.5s,and the misalignment angle is less than 0.5°.Compared with the classic multi-vector in-flight alignment algorithm,the alignment time is reduced by 40% and the alignment accuracy is improved by 80%.Next,aiming at the problem of fine alignment with gross errors in GNSS and combined with the characteristics of the ballistic environment,the pitch and heading information obtained by the ballistic velocity are introduced as the measurement of kalman filter,and the observability of kalman filter in the ballistic environment is analysed.Aiming at the problem of GNSS gross error detection and isolation,the adaptive filtering algorithm based on gross error detection and isolation and the robust filtering algorithm based on prediction residual are described.The simulation results indicated that the observability of the attitude is greatly improved after adding the pitch and heading measurement information.Meanwhile,in the case of gross error in GNSS,both methods can effectively reduce the influence of gross error on the filter.The convergence of misalignment angle of the first method is faster than the second while the convergence of velocity and position error of the first method is slower.Finally,semi-physical experiments are carried out.The first 20 s data segement is selected to verify in-flight coarse alignment,the second 110 s data segement under high dynamic and slightly gross error and the third 95 s data under high dynamic and large gross error are selected to verify the in-flight fine alignment.Results indicated that,as for the in-flight coarse alignment algorithm,the alignment of the horizontal attitude can be completed in 10 s,and the error is less than 0.5°;the alignment of the azimuth can be completed in 15 s,and the error is less than 2°.As for the in-flight fine alignment algorithm,in the case of less GNSS gross errors and without loss of lock,the first method is better than the second;while in the case of large GNSS gross errors and long-term loss of lock,the second method is better than the first.