Research On Initial Alignment Method Of Airborne Strapdown Inertial Navigation System Based On Robust Kalman Filter

Strapdown Inertial Navigation System(SINS)has become the mainstream of airborne inertial navigation system development because of light weight,small size and flexibility.In the airborne laser communication system equipped with SINS,the initial alignment of SINS is the premise of quickly establishing a stable communication link.However,interference such as motion changes and atmospheric disturbances will decrease the accuracy of initial alignment.In order to improve the accuracy and anti-interference ability of the initial alignment and shorten the capture time of airborne laser communication,this paper conducts research on the key technical issues of the initial alignment of SINS under the airborne laser communications.The content is as follows:(1)The principle of airborne SINS and initial alignment is studied.The SINS attitude,velocity and position are calculated and updated by navigation calculation algorithm via the relationship conversion between coordinate systems.The error equations of SINS attitude,velocity and position are deduced respectively according to the error measurement model when the inertial measurement unit is installed.(2)The pointing error,as well as the dynamic and static error of the two-dimensional turntable shaft system are studied.The influence of random error sources on the initial alignment accuracy is studied based on the discussion of the SINS alignment error equation combined with the actual initial alignment mechanism.The equation of analytical coarse alignment is deduced,and the rough attitude matrix is obtained.Based on the above,the fine alignment research is carried out,and then the random error model of the system is established in order to obtain the attitude matrix that meets the accuracy requirements.(3)A fine alignment model based on the Kalman filter for airborne SINS is established,and a robust Kalman filter algorithm is proposed,which is designed by the Riccati equation.This algorithm can reduce the uncertainty of the system.The variance of the filtering estimation error is limited to make the system quadratic stable so that the influence of filtering divergence can be reduced.Analytical coarse alignment simulation is carried out under the static base,and the results show that the alignment error can meet the requirement of small misalignment angle for fine alignment.Aiming at the interference of the airborne laser communication environment,the SINS fine alignment simulation is carried out for comparison of different filters.The results show that the robust Kalman filter has better anti-interference ability with these parameters,and can provide more reliable state estimation,thereby reduce the interference caused by small modeling inaccuracies.(4)Experimental verification.Based on the analysis of the motion of the airborne platform,a simulating flight platform is built,on which the experiments of the initial alignment of SINS in airborne laser communication with different frequencies and amplitudes are carried out.When the frequency and attitude information is concentrated at 0.2 Hz,4°/s,the traditional Kalman filter starts to converge at about 58 s,and the tracking error fluctuates around 1.20-1.40mrad;the robust Kalman filter starts to converge at about 44 s,and the tracking error fluctuates around 1.01-1.11 mrad.Experiments show that when the frequency and attitude information of the simulated aircraft platform is concentrated at 0.2Hz,4°/s,the convergence time of the robust Kalman filter is reduced by 24.1% compared with the traditional Kalman filter,and the spot tracking accuracy is increased by 18.5%.Through the comparison of the initial alignment experiment of different frequency attitude information,it is shown that the robust Kalman filter has stronger antiinterference ability than the traditional Kalman filter,and can perform initial alignment faster;as for the anti-interference ability,the robust Kalman filter has better performance in amplitude than in frequency,which shows that it is more sensitive to amplitude interference than to frequency interference.