Sensors hybridization for distributed launcher system navigation development: first results of the NIBBIO project

Accurate attitude determination is crucial for optimal performance and effectiveness of launch vehicles. While solutions traditionally employed on launchers guarantee high precision, their implementation is also often characterized by extremely high costs associated with “one-off” use in expendable launch vehicles. The recent advancements in the field of low-cost Micro Electronic Mechanical Systems (MEMS) have opened the doors to more economical hybrid navigation schemes, combining measurements from a diverse array of low-cost sensors. Limitations in their use are centered around noise present in their measurements restricting their accuracy and reliability. These limitations may however be overcome through the implementation of filtering and data fusion algorithms to produce more accurate attitude estimates, enabling this sort of solution for future use in space launch vehicles. The Sapienza Space System and Space Surveillance Laboratory (S5Lab) is participating in a research project coordinated by the Italian Space Agency (ASI) for the study and prototyping of hybrid-distributed navigation systems for launchers. In the paper, a simulative model of a three-stage launch vehicle is used to generate noisy measurements from “on-board” sensors. The sensor suite includes measurements from an Inertial Measurement Unit (IMU) made up of a gyroscope and an accelerometer. An Extended Kalman Filter for attitude, position and velocity determination is then developed, and its performance assessed using these same noisy measurements along with measurements obtained from a Global Navigation Satellite System (GNSS) antenna tracking a simulated constellation indicative of those deployed currently. Performance is evaluated by comparing estimates with reference values extracted from the simulative model. A Montecarlo analysis of the filter’s ability in attitude estimation is then performed for varying levels of measurement noise to review filter performance. Finally, the implementation of multiple IMUs is considered, testing the effect of additional sensors on the accuracy of the produced attitude estimates. This work will present preliminary simulative results, which have shown consistent improvements in Euler angle determination when using an array of multiple low cost, low-quality sensors, and precise positioning and velocity estimation through the GNSS antenna. A technology prototype demonstration is further planned for the BEXUS 34 flight within the framework of the RETINA mission, aimed at confirming the simulative findings and investigating the feasibility and performance of the system to increase its Technology Readiness Level (TRL) for future launch vehicle applications.