Measurement Methods Based On Star Sensor Satellite Three-axis Attitude

Attitude control is one main study in the field of spaceflight. The real-time measure of spacecraft attitude is significant to guide, location, especially remote imaging meaure. Using areal array CCD as sensor parts and planets as sensor targets, the star snsor has been high precise attitude sensor components,which board satellite,spacecraft and space station, by means of some advantages, such as high existing,high reliable and high precision, moreover, high autonomy,low weight, low consumed power.In the middle of 1990s', second -generation CCDsensors, CHAMP ASC and AST-201, have been developped, they have some characters, such as high autonomous, large FOV , compensation accomplished inside and direct output. Presently, our nation has achived success in star sensqr main technique, but can't meet high precise imaging measure in FOV, as well as long-time on-orbit working. Under the condition of having no laboratory reports on second-generation sensor system in domestic, several studies are made concntrating on second-generation sensor system in this paper.1.The paper addresses every precision requirements during the design of star sensor on the basis of the relating documents.2.In this paper,employs the Spherical Rectangles Method to store star catalog, and improves it; establishes position transfer function; only usese star diagonal distance as identification character, which avoid the faulse identification caused by using magnitude as character.3.Construct maximum connection match tree identification algorithm with respect to the characters of observed star map and guide stars -star diagonal distance and connection.4.Using BORLAND C++ preformed the whole processing'.forming on-boar storage star catalog with star catalog,coordinate transfer, star identification and attitude acquisition ect.5. Using main frequence 120M computer simulates star sensor, then randomly extract FOV(10 stars in one FOV) from a uniform star catalog that contains 9107 stars, each observed star adds in Gaussian distribution deviation, performs identification and attitude calculation, then analysis the influence of the size of FOV, position noise, threshold, the .number of measured star ect. on identification efficiency, time, as well as the influence of position noise and relating the number of star on the measured attitude accuracy. When the threshold is 0.06? position noise is 5 = 0.02?in the FOV of 12×12 the star identification probability in each region is 100%, the identification time is less than 0.2s, the time of attitude acquisition is less than 0.5s.