Research On Star Identification Algorithm Of All-Time Star Sensor With Small Field Of View

As one of the attitude sensors with the highest measurement accuracy,star sensor can provide accurate attitude information for spacecraft,ships,aircraft and other space and near ground vehicles.Based on the application environment of small field of view and focusing on solving the star identification of all-time star sensor,this paper mainly studies the construction of star sensor guide star catalog and typical star identification algorithm of all-time star sensor with small field of view.The contents are as follows:1.The guide star catalog of all-time star sensor with small field of view based on2 MASS star catalog is constructed: according to the all-time application scenario,the advantages of infrared star detection and the characteristics of daytime star detection environment are studied,and the near-infrared band is selected as the working band of star sensor;After the qualified navigation stars are selected from the 2MASS star catalog,the navigation star catalog is homogenized to improve the efficiency of star identification.2.The recognition performance of triangle and pyramid recognition algorithms is studied and compared in the all-time application environment with small field of view.Through theoretical derivation,the imaging model of star sensor and the influence of error parameters on Star angular distance are deeply analyzed;the classical star identification algorithms triangle and pyramid algorithms are simulated and experimentally analyzed.The results show that the star identification algorithm based on angular distance matching is easy to be affected by star position error and focal length deviation,and triangle and pyramid algorithms can not meet the requirements of small field of view in all-time star identification.3.A star identification algorithm based on attitude information is proposed.According to the known optical axis direction,the stars in the field of view in the guide star catalog are quickly retrieved.The information is used to assist the star identification of triangle algorithm to solve the problems of reduced recognition rate and reduced recognition efficiency of triangle and pyramid algorithm when the star extraction error is large.Simulation and experimental results show that the recognition performance of this algorithm is much better than traditional algorithms when the star angular distance threshold error is large.