On-orbit calibration based on second-order nonlinear filtering

This paper presents a real-time on-orbit calibration algorithm for gyros and star sensor with second-order nonlinear filtering to meet the high accuracy of satellite three-axis attitude determination. A model of gyros misalignment, scale factor error and drift is constructed by considering the error sources during satellite attitude measure. Compared with the other calibration algorithms, this algorithm provides more attitude information for satellite attitude determination and correction and makes sure the long-term on-orbit working precision of attitude sensors. Finally, a simulation is carried out with the data from the on-orbit satellite ldquoTS-1rdquo of Harbin Institute of Technology (HIT). The results are identical with the actual results and indicate the availability and feasibility of the on-orbit calibration algorithm.