| The essense of in-orbit calibration on attitude sensors is that it can reduce the sysyemerror e?ectively. With the increasing demand of precision, in-orbit calibration will be ap-plied more and more widely. This paper takes the attitude determination system composedof star tracker and gyroscopes as research object, focuses on two major aspects:researchon the factors in?uencing the results from star tracker, calibration model is bulit and ef-fective method is proposed; research on the alignment among attitude sensors, definitebase sensor is given, payload is added, and then several estimator is presented for thecalibration, in the end a reducing dimensions calibration scheme is proposed.1.Traditional used measure variance is based on the simplified by the assumptionof small FOV. And with the development of star tracker, larger FOV is more and morecommon. In this situation, a new, fine variance model is needed to be studied. This paperpresents two theoretical models, one is called light cone variance and the other is calleddistortion variance.2.The drift of imaging parameters of star tracker could a?ects the results of attitudedetermination, this paper presents a method to calibrate the principle points drift, focaldrift, distortion of star lens using equal angular criterion. A genetic algorithm is used tosolve the problem.3.When star tracker rotates fast, the precision of the star tracker output will be a lagsignal, and the magnitude of measure variance will get larger. A model is proposed, anda calibration method is presented.4.Alignment calibration on the system composed of star tracker, gyroscopes andoptic payload is researched. Since the payload is added to the calibration. Obsolute cali-bration can be proceeding. Several estimators are given, and the comparation is analyzed.5.For the high dimension of calibration system this paper proposed, a reduced di-mension calibraiton system based on the analysis is presented. |
PDF Full Text Request