Autonomous satellite navigation and orbit-keeping using attitude sensors

The design concept of an autonomous navigation and orbit-maintenance control architecture for earth-observation satellites is presented, and results of a Monte Carlo simulation are reported. The proposed system uses attitude-sensor data (from star tracker or fine sun sensor) to optimize solar illumination angle and ground trace and is implemented using an onboard microcomputer. Conditional-Gaussian filtering theory is applied to a discrete linearized model of the system, and the unbiased minimum variance of the recursive estimation of orbital variables in the presence of orbit control and sensor tracking is confirmed. An adaptive filtering algorithm is developed to reduce the effects of attitude-determination error. The results of a Monte Carlo simulation using typical orbital and error values are presented in graphs and tables and shown to indicate successful performance of the required control operations.