Previous Abstract Return to Session B8 Next Abstract

Session B8: Quantum PNT 3

Gradiometry-Based Sensor Fusion with Fault Detection and Isolation for Autonomous Cislunar Navigation
Jillian Ross and Rami S Mangoubi, C.S. Draper Laboratory
Location: Ballroom E
Date/Time: Tuesday, Jun. 2, 4:25 p.m.

Motivation and Background. Cislunar missions are of long duration, even multi-year, and unaided robust autonomous navigation accuracy is essential to achieving safety requirements for Artemis-relevant mission profiles, such as the NASA Lunar Gateway Near Rectilinear Halo Orbit (NRHO) and Orion Trans-Earth Injection (TEI). Both accuracy and reliability depend on judicious sensor specification selection and/or fusion of their outputs, and on sensor redundancy.
Previous work. An earlier study (Ibars et al.) focused on quantum gravity gradiometers alone and demonstrated that a milli-Eotvos noise-level instrument satisfies the Gateway mission requirements. Another study (Ross and Frueh) considers the neutron star pulsar alone and demonstrates that it satisfies the requirements for safe Earth entry in the Orion TEI and safe maneuvering over a 21-day period in a Gateway-inspired NRHO.
Contributions. The present study builds upon previous work by considering gravity gradiometers, magnetic gradiometers, and neutron star pulsars, and presents early results for sensor specifications that, when used in an Extended Kalman Filter based on the Circular Restricted Three-body Problem or CR3BP simulation with process noise added, meet both NASA-specified mission requirements for cislunar autonomous navigation, either by using individual sensors, or by fusing multiple ones. Specifically, position errors as low as 2 m can be reached for the Gateway mission, and flight path angle errors at Earth Interface as low as 0.005 degrees for the Orion Moon-to-Earth transfer. The study will illustrate the ability to deal with sensor failure or temporary blackouts of a single sensor.
Methodology. To simulate the vehicle cislunar trajectory, the study relies on the CR3BP dynamics simulation. Process noise is also added to account for deviation from expected trajectory, as well as other disturbances such as solar radiation pressure.
The three sensors together provide fifteen measurements: the gravity gradiometer, consisting of two accelerometers per axis, produces a 3x3 symmetric tensor measurement of the Jacobian of the earth’s gravity and moon’s gravity vectors superimposed, or equivalently, the Hessian of the combined earth’s and moon’s gravity potential. The magnetic gradiometer, consisting of two magnetometers per axis, outputs a 3x3 symmetric tensor measurement of the Jacobian of Earth’s magnetic field potential. Finally, the x-ray pulsar sensor measures three pulsar signals for full positional information according to occultation availability, delivering a 3x1 measurement vector where each element is a range between the spacecraft and Earth.
White noise alone, as well as white noise with bias and scale factors, are added to each sensor in order to compare how different combinations of sensor measurement errors are traded off while satisfying the two cislunar mission requirements, namely Gateway and Orion.
Preliminary Results: Gateway NRHO mission. The results indicate that the approach satisfies NASA’s mission requirement, using a GOCE mission gravitational gradiometer with sensitivity of 0.01 to 0.02 Eotvos/sqrt-Hz, with a very large initial error condition or 10,000 km. Specifically, an error of 10 km or less as required is reached within 10 days. If a quantum gravitational gradiometer of 0.001 Eotvos error standard deviation is used, such as the optical Lattice Atom Interferometer, OLAI, the mission requirement is satisfied within 5 days. If a magnetic gradiometer of 1 nano Tesla per meter is added, the sensor fusion yields less than 10 km error within four days only. In addition, the sensor redundancy endows the navigation system with increased robustness to sensor faults that may occur.
Preliminary Results: Orion Moon-to-Earth Transfer. Using sensors fusion as for the Gateway mission, NASA’s autonomous navigation requirements to ensure safe Earth entry are also satisfied. Specifically, the gravity and magnetic gradiometer sensor fusion configurations demonstrate convergence in flight-path angle standard deviation mapped to Earth Interface to the required value of .5 degree (3-sigma).
Equations, figures, and references will also be presented along with additional results that include more general sensor model errors, such as bias.

Previous Abstract Return to Session B8 Next Abstract