Architectural Elements of Hybrid Navigation Systems for Future Space Transportation

Abstract

In this work we analyse several architectural elements and design aspects of a hybrid GNSS/INS navigation system conceived for space transportation. The fundamental limitations of the purely-inertial method, currently employed by most launchers, has raised interest for GNSS-aided solutions. Combination of inertial measurements and GNSS outputs allows inertial calibration on-line, solving the issue of inertial drift. However, many challenges and design options unfold. The most fundamental architectural features such as coupling depth, modularity between filter and inertial propagation, and open-/closed-loop nature of the configuration, are discussed in the light of the envisaged application. Importance of the inertial propagation algorithm and sensor class in the overall system is investigated, being the handling of sensor errors and uncertainties that arise with lower grade sensory also considered. In terms of GNSS outputs we consider receiver solutions (Position and Velocity) and raw measurements (Pseudorange, Pseudorange-rate and Time-Differenced Carrier Phase). Receiver clock handling options and atmospheric error correction schemes for these measurements are analysed under flight conditions. System performance with different GNSS measurements is estimated through covariance analysis, being the differences between Loose and Tight coupling emphasised through partial outage simulation. Finally, we discuss options for filter algorithm robustness against non-linearities and system/measurement errors. A possible scheme for Fault Detection, Isolation and Recovery is also proposed.