Diversity Architecture for Robust GNSS/INS Navigation in Launcher Applications

Launcher’s navigation systems have traditionally relied on high-grade INS (Inertial Navigation Systems) to inject a payload in the desired orbit. As launcher operations become more frequent and complex, even involving the automatic landing of re-entry stages, GNSS (Global Navigation Satellite System) positions itself as an enabler technology to guarantee the success of the operations, not without its challenges to solve. The main objective of this work is the design of a hybrid GNSS/INS navigation system and FDIR (Fault Detection Identification and Recovery) algorithms used to demonstrate the robustness against errors in GNSS and INS technology. The navigation solution is provided by a modular sensor fusion algorithm architecture, which combines inertial, GNSS, radar-altimeter and star sensor measurements to satisfy the accuracy requirements for all the flight phases. Indeed, the architecture reflects the need to adapt to multiple launcher configurations such as expendable launch vehicle (Vega, Ariane-5), micro launchers (Shefex-2), reusable first stage boosters (Falcon-9) and unmanned re-entry vehicles (Space Rider), in which the most critical phases of the flight have been considered for the study. The performance of the navigation system is assessed both in ideal conditions and under meaningful threats/failures which aid in the development of the FDIR algorithms. The threats include GNSS signal outages/loss of tracking, satellite/receiver clock bias/drift discontinuities, spoofing, receiver hardware failures, IMU saturation, vibration rectification, coning & sculling effect, and INS software numerical failures. To that purpose, we developed a simulation environment consisting of a RFCS (Radio Frequency Constellation Simulator), a Qascom QN400 space receiver and MATLAB software, which in turn enables the simulation of the navigation sensors, the desired threats/failures, and the validation of the navigation system performance. The simulation environment also allows the selection of different inertial sensor models or the usage of a real GNSS receiver.