Fully Autonomous Layered Star Identification Algorithm Research Based-on Star Sensor

The thesis is a research work based-on star sensor about star identification algorithm, construct of guide star database, attitude calculation method.The thesis first simply introduces the satellite attitude control system and some attitude sensors. Introduces the star sensor's function, component and working theory, development of star sensor and its main technique. Analyzes the technical characters of star sensor that influence star identification, gives the requirement of software about star identification.The thesis develops new techniques on the basis of the relatingdocuments--Construct of Reduced 2-D Indexed Layered Guide StarDatabase. The database is a uniform storage reduced guide star database with improved 2- D index It divides guide stars to local bright star (LBS) and local dim star (LDS), and gives priority to LBS in making star triangle to reduce the number of stored triangles. It uses 2-D Na index and linear storage to improve the hit the target rate, save the memory for index. Expands the index at the edge of Na to reduce the influence of position error.The thesis develops new techniques on the basis of the relating documents ?Fully Autonomous Layered Star Identification Algorithm. Fully Autonomous Layered Star Identification Algorithm, which divides the detected stars to LBS and LDS, gives priority to LBS in making star triangle and identification, so can find the matching star quickly. It takes the maximum matching reliability and expanding index at the edge of Na to reduce the influence of position error, magnitude error and fake star, sohave high identification correct rate. It can be used for full and local sky star identification.Least squares solution of equation is used to calculate and adjustment transformation matrix at same time, transformation matrix is used to calculate the 3-axis angle of star sensor. So is very fast and precise.Above methods are tested and analyzed by computer simulation. Computer that is a PHI with main frequency 650M and windows as operating system simulates a new generation star sensor which diameter of circle FOV is 20arcdeg, position uncertainty is lOarcsec, magnitude uncertainty is 0.2mag, one fake star. Performs 10,000 experiments of fully autonomous star identification and attitude calculation, the result is: guide star database is 2. 3MB, full sky cover rate is 99. 92%. Identified star correct rate is 99. 87%, identified FOV correct rate is 98. 59%, mean time of star identification and attitude calculation for one experiment is 1. 3ms, yaw angle mean square deviation is 6. 8arcsec, pitch angle mean square deviation is 6. Sarcsec, rotation angle mean square deviation is 52. 3arcsec.