FAST THREE-AXIS ATTITUDE DETERMINATION USING VECTOR OBSERVATIONS AND INVERSE ITERATION.

AN ALGORITHM IS DEVELOPED TO FIND THE OPTIMAL ROTATION MATRIX THAT TRANSFORMS A SET OF REFERENCE VECTORS INTO A CORRESPONDING SET OF MEASUREMENT VECTORS. THE ALGORITHM MINIMIZES AQUADRATIC LOSS FUNCTION IN WHICH THE ROTATION MATRIX IS REPRESENTED AS A QUATERNION. THE MINIMIZING QUATERNION IS SHOWNTO BE AN EIGENVECTOR OF A SPECIAL MATRIX. THIS EIGENVECTOR IS FOUND BY INVERSE ITERATION. THE ALGORITHM IS COMPUTATIONALLY EFFICIENT AND ELIMINATES A SINGULARITY CONDITION CONTAINED IN A PREDECESSOR ALGORITHM.