Extension of Farrenkopf Steady-State Solutions with Estimated Angular Rate

摘要

Steady-state solutions of the single-axis attitude estimation filter are considered for the traditional case of dynamic model replacement, where the angular rate is determined directly from a gyroscope reading and bias estimate, and for the augmented case where the angular rate is estimated in the filter. The impact of the process noise on the steady-state values is analyzed, and it is found that there exists a value for the process noise below which the augmented filter performs better, and above which the dynamic model replacement filter performs better. A Multiple-Model Adaptive Estimation filter is developed to allow for uncertainties in assumed process noise in the augmented filter to be handled. Three-axis results are compared to single-axis results and are found to perform similarly.