Study of attitude determination accuracy and Earth-point stabilization performance using asynchronous star tracker and angular velocity sensor measurements
- Paper number
IAC-24,C1,1,6,x88981
- Author
Dr. Danil Ivanov, Keldysh Institute of Applied Mathematics, RAS, Russian Federation
- Coauthor
Dr. Yaroslav Mashtakov, Keldysh Institute of Applied Mathematics of RAS, Russian Federation
- Coauthor
Ms. Uliana Monakhova, Keldysh Institute of Applied Mathematics of RAS, Russian Federation
- Year
2024
- Abstract
A problem of satellite attitude real-time determination using angular velocity sensor and star tracker is considered in the paper. Although this problem is well studied in the literature, there are still several issues to be dealt with. First of all, it is asynchronous measurements, when information on angular velocity and attitude quaternion is available onboard at different times and with different frequencies. In addition, attitude and angular velocity estimations are required at the times of control inputs calculation, which in general does not match the time of any measurements. The second problem is a dependence of star tracker measurements accuracy on the satellite angular velocity: at higher angular rates accuracy deteriorates, and attitude measurements might even become unavailable. This problem is especially crucial for remote sensing satellites: when camera optical axis tracks the point of interest on Earth surface, the angular velocity greatly increases during the flyby. The third problem is angular velocity sensor bias: it may significantly drifts over time, and it must be corrected in real-time. The paper proposes two attitude determination algorithms based on extended Kalman filter, which able to deal with all of the aforementioned problems. First algorithm uses kinematic relations, it estimates only the angular velocity sensor bias and attitude quaternion. It does not require information on satellite parameters, it is characterized by less computational burden, though the angular velocity estimation accuracy is limited by standard deviation of the sensor random noise. The second algorithm is based on both kinematic and dynamic motion equations: it estimates angular velocity sensor bias, angular velocity and attitude quaternion. The satellite tensor of inertia, reaction wheels parameters and history of control inputs are required for state vector estimations. The performance of these algorithms is compared under different attitude motion scenarios and disturbances, taking into account different thresholds for star tracker measurement unavailability, uncertainties in tensor of inertia, and delays in measurements time stamps. The influence of the state vector estimation errors on the Earth-point stabilization accuracy is studied.
- Abstract document
- Manuscript document
IAC-24,C1,1,6,x88981.pdf (🔒 authorized access only).
To get the manuscript, please contact IAF Secretariat.