Michael Lichter, Air Force Institute of Technology and NASA Glenn Research Center
Keywords: Star Tracker, Communications, Optics, Model, Navigation, Attitude
Abstract:
Improvements to spacecraft attitude estimation make possible beaconless laser communications. Current laser communication systems require an active beacon from the receiver so that the transmitter can point at the receiver and maintain the link. For deep space mission applications, this becomes impractical as the beam power required becomes too large. High precision attitude measurement systems obviate the need for the beacon from the receiver making it possible for the spacecraft to beam a laser communications signal to a ground station without the ground station advertising its location.
The presentation will discuss research to target this improvement which includes developing new detection and estimation methods to improve the accuracy in locating stars on a focal plane detector, developing an understanding of the effects of changes in the optics design parameters and aberration, including defocus, on the navigation solution itself. This understanding can lead to an optimization of the attitude solution with respect to those optics realm parameter changes.
The methodology discussed includes the development of a model of a current star tracker system in Matlab to be used to simulate the realistic physical behavior of the system, including the star inputs, noise, and the navigation output solution. Using this model, multiple algorithms are implemented, including a multi-hypothesis method (MHT), to detect and estimate the position of the stars on the focal plane detector. A translation calculation to derive the position in the celestial sphere from that estimate is used. It will be shown that using an MHT detection and estimation algorithm developed at the Air Force Institute of Technology (AFIT) for detection and estimation, a greater accuracy can be found for each star estimation relative to more traditional detection and estimation algorithms. The approach then uses the model to develop statistics of the star tracker and the attitude estimation outputs to understand the accuracy, or variance, of the systems attitudes solution. This solution is repeated for a range of defocus aberration, and a lower limit to the variance of the attitude solution is shown. The effect of under-sampling is also taken into account for this analysis.
Using this approach, one can simulate a star not as a Gaussian spot on the focal plane as done in previous work, but use of a spot that includes the effects of aberrations of the optic system, and the effects of under-sampling and noise from the focal plane detector as well.
To validate the model and statistical analysis, a Cramer Rao lower bound solution is derived for the star tracker system. It is solved for the same range of defocus aberration as used for the Monte Carlo Analysis. The results for the two are compared and shown to correlate very well.
The research provides insight into an attitude solution design process that allows the designer to understand the effects of changes in the optics design realm to the navigation solution itself, or the star attitude output of the star tracker, and use a number of parameters to optimize the accuracy of that solution. With these tools, the star tracker design can hopefully approach accuracies that have not been available to this point. A goal of a few urads overall pointing requirement for a deep space optical communications asset may is desired, and can be aided with the help of these design analysis tools and insights. Analysis includes investigating if a star tracker’s accuracy can be improved by possibly removing focus error and allowing more under-sampling effects, possibly contradicting conventional wisdom and approaches.
Date of Conference: September 17-20, 2019
Track: Space-Based Assets