Precise Real-Time Relative Orbit Determination for Large-Baseline Formations Using GNSS

Abstract

This paper presents a real-time GNSS-based navigation architecture for distributed space systems, which performs precise relative orbit determination at separations of hundreds of kilometers. Through differential carrier-phase GNSS and on-board integer ambiguity resolution, a hybrid Kalman filter combines the extended Kalman filter time update with the unscented Kalman filter measurement update to exploit nonlinearities in separation-dependent errors while minimizing computational load. Differential ionospheric path delays on multiple frequencies are modeled through the Klobuchar model with an estimated correction term to account for unmodeled effects. This method allows the filter to use low-noise single-difference carrier-phase measurements at large separations to provide relative navigation solutions with centimeter-level error.