Improvement of the Long-Term Orbit Prediction for LEO Navigation Satellites Using the Inner Formation Method

In this paper, the concept of the inner formation navigation satellite is proposed. It operates in the low Earth orbit with very high autonomy to make the requirement for ground support minimized. A state transformation matrix-based orbit fitting method is presented for orbit prediction, and the long-term accumulation of prediction errors is investigated by simulations. The fuel consumption for orbit maintaining with a linear controller is discussed. Results show that the orbit can be predicted autonomously to the meter-level accuracy for 90 days if the constant component of residual nongravitational disturbance can be suppressed to 1 x 10(-13) m/s(2). The maximum fuel consumption for 5 years is on the order of 10 kg if optical sensors and electric thrusters are used for orbit maintaining.