A Novel Algorithm on Sub-meter Level Real-Time Orbit Determination Using Space-Borne GPS Pseudo-Range Measurements

Abstract

The real-time onboard orbit determination using GPS measurements has been rapidly developed in recent years due to its global coverage, abundant observations and low-cost. It can provide accurate orbital parameters for many space tasks such as LEO satellite orbit control and earth observation. The position and velocity accuracy of traditional real-time orbit determination algorithm using space-borne GPS pseudo-range measurements are always up to 1.0 m and 1.0 mm/s (3DRMS) because of an inevitable limitation of error of GPS broadcast ephemeris, thus it is difficult to meet the real-time high-precision requirements of High-resolution Earth Observations System and other space missions. Through thoughtful analysis of the variation of broadcast orbit and clock offset errors, a novel real-time orbit determination method using GPS pseudo-range observations was present in this paper. The new method estimates corresponding parameters in the Kalman filtering algorithm to absorb the slowly varying errors of broadcast orbits and clock offsets, so as to achieve orbit results with accuracies of sub-meter level in position and sub mm/s in velocity. Then a simulative test is carried out to process the space-borne GPS dual frequency pseudo-range data of consecutive 31 days from GRACE-A satellite using the auto-developed software SATODS. The test demonstrates that the position and velocity errors (3DRMS) of orbit results of the new method are reduced to 0.4–0.6 m and 0.4–0.6 mm/s respectively, which means orbital accuracies are improved by more than 40 % compared to the traditional one. Additionally, the new method has the same strategies in dynamic model and data pre-processing as the traditional one, so it would not increase computational burden visibly and has a very strong practical value.