Attitude-determination filtering via extended quaternion estimation

The quaternion estimation (QUEST) batch attitude-determination algorithm ha s been extended to work in a general Kalman-filter framework. This has been done to allow the inclusion of a complicated dynamics model and to allow t he estimation of additional quantities beyond the attitude quaternion. The QUEST algorithm, which works with vector attitude observations, serves as a starting point because it is able to work with a poor (or no) first guess of the attitude. This paper's extended version of QUEST uses square-root in formation filtering techniques and linearization of the dynamics to propaga te the state and its covariance. The measurement update problem is solved b y a technique that is an extension of the original QUEST algorithm's eigenv alue/eigenvector solution. The paper demonstrates the new algorithm's perfo rmance on an attitude determination problem that uses star-tracker and rate -gyro measurements. The new algorithm is able to converge from initial atti tude errors of 180 deg and initial rate-gyro bias errors as large as 2400 d eg/h.