All spacecraft attitude estimation methods are based on Wahbas optimization problem. This problem can be reduced to finding the largest eigenvalue and the corresponding eigenvector for Davenports $K$-matrix. Several iterative algorithms, such as QUES
T and FOMA, were proposed, aiming at reducing the computational cost. But their computational time is unpredictable because the iteration number is not fixed and the solution is not accurate in theory. Recently, an analytical solution, ESOQ was suggested. The advantages of analytical solutions are that their computational time is fixed and the solution should be accurate in theory if there is no numerical error. In this paper, we propose a different analytical solution to the Wahbas problem. We use simple and easy to be verified examples to show that this method is numerically more stable than ESOQ, potentially faster than QUEST and FOMA. We also use extensive simulation test to support this claim.
