GYROLESS BACKUP ATTITUDE ESTIMATION ALGORITHM FOR NANOSATELLITES USING MAGNETIC FIELD DERIVATIVES IN THE LOSS OF SUN SENSOR MEASUREMENTS

Sirin Yakupoglu, Mehmet Esit, Chingiz Hajiyev and Halil Ersin Soken

The main aim of this study is to design a low-cost attitude lter for nanosatellites. Since Sun sensors and magnetometers are widely used commercial off-the-shelf attitude sensors, the attitude estimation algorithms using these sensors are needed for either primary or backup modes. Throughout the study, QUEST aided multiplicative extended Kalman lter (MEKF) is used to obtain ne attitude and angular velocity estimation. Although this set of sensors has proven to give satisfactory estimation results, the problem arises when Sun is not visible. For the eclipse period, the estimation must be updated by only magnetometer data. Magnetometer measurements are used to generate another vector pair based on magnetic eld derivative and QUEST processes these two vector sets to provide attitude information for updating the MEKF. The proposed method provides 5 degree attitude accuracy in eclipse period without changing the structure of the lter between different conditions. When Sun sensor is available, ne estimation is achieved as having lower than 1 degree attitude error.

Fulltext (PDF)