GPS-BASED AUTONOMOUS SPACE NAVIGATION

COMMERCIAL AND GOVERNMENT SPACE TRANSPORTATION SYSTEMS DEPEND ON GROUND SUPPORT FOR NAVIGATION. THE GROUND SOLUTION CAN BE COSTLY AND MAY PROVIDE LESS ACCURACY THAN NEEDED. A GPS-BASED REAL-TIME AUTONOMOUS SPACECRAFT NAVIGATION SYSTEM IS PROPOSED. AN ALGORITHM WILL BE DEVELOPED WHICH COMBINES GPS-DERIVED POINT POSITION ESTIMATES WITH A PROPAGATED A PRIORI POSITION ESTIMATE. THE POINT POSITIONS ARE COMPUTED BY DIRECT ALGEBRAIC SOLUTION OF THE PSEUDORANGE EQUATIONS WITHOUT RESORTING TO LINEARIZATION. THE MOST APPROPRIATE METHOD FOR PROPAGATION OF THE A PRIORI ESTIMATE WILL BE DETERMINED FROM METHODS SUCH AS MONTE CARLO ANALYSIS, INTEGRATION OF PARTIAL DIFFERENTIAL EQUATIONS, AND STATE TRANSITION EQUATIONS DERIVED FROM THE ORBITAL EQUATION OF MOTION. A BAYESIAN UPDATE WILL BE USED TO COMBINE THE GPS POINT ESTIMATE WITH THE A PRIORI ESTIMATE. THE MAXIMUM A POSTERIORI VALUE OF THE POSITION DENSITY FUNCTION WILL DEFINE THE BEST VEHICLE POSITION ESTIMATE. THE ALGORITHM WILL BE IMPLEMENTED ON A DESK TOP COMPUTER AND TESTED FOR ACCURACY, EASE OF IMPLEMENTATION, AND COMPUTATIONAL BURDEN. THE DELIVERABLES WILL INCLUDE THE COMPUTATIONAL ALGORITHM, THE COMPUTER CODE, THE TEST RESULTS, AND A PLAN FOR PHASE II IMPLEMENTATION OF THE ALGORITHM ON PROTOTYPE HARDWARE.