Adaptive Sequential Estimation with Unknown Noise Statistics.
Math Lab preprint series no. 7 (Final),
AEROSPACE RESEARCH LABS WRIGHT-PATTERSON AFB OHIO
Pagination or Media Count:
Sequential algorithms are derived for suboptimal adaptive estimation of the unknown state and observation noise statistics simultaneously with the system state. First- and second-order moments of the noise processes are estimated based on noise samples generated from quantities in the usual Kalman filter algorithm. A limited memory formulation is developed for adaptive correction of the a priori statistics which are intended to compensate for time-varying model errors. The new estimators are applied to an orbit determination problem for a near-earth satellite with significant modeling errors. Results indicate that improved state estimates can be obtained at little computational expense when erroneous a priori noise statistics are adaptively corrected in the filter algorithm.
- Statistics and Probability
- Spacecraft Trajectories and Reentry