PUBLISHED PAPERS #4.02
| Demet Cilden-Guler, Chingiz Hajiyev Kinematic Relations Based Nanosatellite Attitude Estimation and Rate Gyros Calibration |
|---|
| Abstract. In this study, the attitude angles of a nanosatellite are estimated using vector measurements from magnetometers and sun sensors. The estimation approach relies on a nontraditional filtering technique that exclusively uses a kinematic model, which is propagated by the measurements from rate gyroscopes. To address the challenge of gyro drift, bias terms are incorporated into the state vector, allowing for the estimation of these biases. The process is carried out in two distinct stages. In the first stage, singular value decomposition (SVD) is employed to determine the attitude measurements. In the second stage, an extended Kalman filter (EKF) is developed, which operates on the linear attitude measurements. These two stages are integrated into a single comprehensive algorithm, which accurately estimates both the attitude angles and gyroscope biases. To improve the accuracy of the rate gyroscope data, the gyroscope bias estimates are subtracted from the rate gyroscope measurements at each estimation step, effectively correcting the measurements. Additionally, the study highlights the effectiveness of the combined approach, where the nontraditional filtering technique provides robust performance even in the presence of noise or other errors in the system. By continuously refining the gyroscope bias estimates, the method ensures that the attitude estimation remains highly accurate over time, contributing to the overall stability and reliability of the nanosatellite's orientation system. |
| Keywords: nanosatellite, attitude estimation, attitude kinematics, Kalman filter, magnetometer, sun sensor, rate gyro |
 Download PDF |
| DOI: |