AAS 96-181

FLIGHT DATA RESULTS OF ESTIMATE FUSION FOR SPACECRAFT RENDEZVOUS NAVIGATION FROM SHUTTLE MISSION STS-69

J. R. Carpenter, NASA Johnson Space Center; R. H. Bishop, The University of Texas at Austin

Abstract

As more and more spacecraft are now being equipped with Standard Positioning Service Global Positioning System (GPS) receivers, the possibility exists for increasingly autonomous rendezous navigation. Although relative state estimation accuracy takes precedence in rendezvous, the increased accuracy of GPS-derived inertial states will more accurately locate the vehicles with respect to variations in the gravity potential, thereby making possible more accurate predictions of their future trajectories. With such information, more accurate maneuvers could be targeted, and fewer correction burns scheduled, saving consumables and freeing up timelines. For the many cases when the target vehicle will have no GPS receiver, or no means of communicating the GPS data even if present to the Orbiter, fusion of the GPS and rendezvous radar data could provide accuracies approaching that of relative GPS. While this fusion could be accomplished with a standard, centralized Kalman filter processing both raw data types, in the case of the Space Shuttle Orbiter, such a filter would largely duplicate the effort expended by the existing Kalman filters resident in the avionics system and the GPS receiver/processor. Fusing the state estimates of these filters instead, as advocated in this paper, is perhaps a better solution. The paper is devoted to showing, using actual Space Shuttle flight data from STS-69, that fusion of the state estimates is a practical and viable approach to achieving relative GPS accuracies with only the Orbiter GPS and Orbiter rendezvous radar.