Attitude Determination Of Star Sensor Based On Spinor

The star sensor provides high-precision attitude information for satellites in space relative to the inertial reference system.Typical static attitude determination algorithms include numerical solutions based on Newton’s iteration,q method and QUEST method,all of which can be reduced to Wahba problem.The cost function is minimized by solving the optimal orthogonal cosine matrix with determinant +1.Because the star is in the infinite celestial plane,there is only direction information and lack of depth information,so the above algorithm is difficult to overcome the uncertainty of the star observation data,and then makes the observation data complex colinearity serious,resulting in low satellite attitude determination accuracy in recent years.In terms of precise kinematics and dynamics modeling of robots,the attitude and position of rigid bodies in space are described by means of spin theory,and the nonlinear characteristics of uncertain systems are modeled.Aiming at the problems of static star-sensitive attitude determination,this paper draws on the method of spin in the analysis of the state of motion of rigid bodies,and proposes a new method of star-sensitive attitude determination.The spin is used to represent the star observation data,and then the satellite attitude is determined.And innovation points are as follows:(1)In the stellar geometric imaging observation,the spin is used to indicate the line of the star point.If the spin is rotated around the imaging visual axis,translation along the imaging direction is bound to occur.The translation distance is the rotation distance.And because the six-dimensional coordinates of the spin do not satisfy the second-order constraint,the spin can be increased along the imaging direction,so that based on the two-dimensional stellar imaging direction,the control conditions of the depth information along the boresight direction are added.Experimental results show that the method in this paper can avoid the problem of large changes in attitude solution caused by small errors in ill-conditioned star observation data,thereby ensuring the stability of attitude solution,and solving the ill-posed problem of stellar observation data system.(2)The method of determining the attitude of the star-sensitive period based on the combination of point and line characteristics is studied.Through the collinear condition equation based on the point feature and the coplanar condition equation based on the rotation,the joint adjustment is solved,and the number and distribution of stars are analyzed.And stellar extraction errors.Impact on satellite attitude determination.Simulation experiments show that when the control conditions are 4 points and2 lines evenly distributed around the star map,the algorithm in this paper achieves the highest accuracy.Compared with the MLS algorithm,the accuracy of the yaw angle is improved,and the overall accuracy is better than the MLS algorithm.Reached 1 arc second.At the same time,under the condition that the star point extraction error is large,the method in this paper can also obtain higher accuracy.