Magnetometer Based Navigation

Magnetometer based navigation (MAGNAV) provides low cost, autonomous navigation for low Earth orbit (LEO) missions. The magnetometer has four primary advantages. First, it is always part of the sensor complement for LEO missions, primarily for momentum management. Second, it always outputs data; that is, it is not subject to occultation or tracking problems. Third, it is very reliable. Lastly, it provides information on spacecraft attitude, rate, and orbit. The system developed in GNCC is based on an Extended Kalman Filter algorithm, combined with a pseudo-linear Kalman filter, producing the full set of navigation parameters, namely attitude, orbit, and rate. Reducing the complexity of onboard processing, eliminating costly sensors, and reducing ground operating costs, while providing accuracy and reliability are additional objectives of MAGNAV.

Typically, the MAGNAV algorithm, in order to provide simultaneous attitude, orbit, and rate estimates, also processes data from an additional sensor, such as a gyro, sun sensor, or GPS (operating alone the magnetometer can provide either attitude and rate or orbit estimates). This improves the accuracy and speed of convergence, and ensures robustness. A magnetometer-gyro configuration has been tested with real data from four GSFC satellites. A magnetometer-sun sensor configuration has been tested with TRACE data and is scheduled for an inflight test onboard the WIRE spacecraft. The magnetometer-GPS configuration (GPSMAG) underwent analytical testing in FY01, with the goal of developing a ‘black-box’ spacecraft navigation system, as depicted in figure 4-10. Example results are given in table 4-1 below. It is expected that MAGNAV could be used in a backup mode; startup mode, e.g. initialization; anomaly resolution; or as a prime navigation system for a LEO mission with coarse requirements.

In-Flight Experiment of MAGNAV on WIRE Spacecraft

An in-flight experiment of the MAGNAV algorithm will take place onboard the WIRE spacecraft. The flight code has been prepared and is undergoing final ground testing. The code is expected to be uplinked and patched into the WIRE onboard computer in early October of 2001, with two weeks of testing to follow. MAGNAV will run as an independent task, in parallel with the fine and coarse onboard attitude determination systems. It is anticipated that this test will demonstrate the capabilities of the MAGNAV algorithm to provide low cost, autonomous estimates of orbit, attitude, and rate for low Earth orbit satellites.

IR&D Funded Research of GPS/Magnetometer Navigation

Analytic testing of the GPSMAG, a version of MAGNAV that incorporates GPS measurements, was conducted during FY01. The testing was performed using Matlab. The spacecraft simulation was based on a UARS ephemeris and included simulation of the GPS constellation. The algorithm was successful in estimating the spacecraft orbit, attitude, rate, and GPS clock errors using simulated measurements from two GPS satellites (both phase and pseudo-range), along with magnetometer measurements. Starting with initial errors of 500 km/axis in position, 0.5 km/sec/axis in velocity, 103 degrees in attitude, and 5 deg/sec/axis in rate, the average RSS errors after 12 hours were less than 0.3 deg in attitude, 0.003 deg/sec in rate, 30 meters in position, and 6.5 cm/sec in velocity. Most of the convergence occurred within the first 50 minutes. Additionally, the algorithm was able to follow a simulated 80-degree rotation about a spacecraft body axis. The results of the testing were presented in two conference papers.

[Technical contacts: Julie Thienel/Rick Harman]