Real-time Onboard Orbit Determination Using GPS Navigation Solutions

Real-time onboard orbit determination utilizing GNSS is a promising approach with broader usage during the last decade. In this paper, we propose a novel real-time orbit determination algorithm based on reduced-dynamic technique using GPS navigation solutions. The proposed algorithm employs Extended Kalman Filter and introduces colored-noise whitening and self-adaptive method to enhance the robustness. Simulation results using real GPS navigation data collected from the CHAMP satellite show that the overall position and velocity errors of the estimated orbit are 10.860m and 0.0128m/s RMS respectively. Furthermore, prototype software of the algorithm has been developed and successfully demonstrated in DSP TMS320C6713.