Improved navigation accuracy through Kalman filtering

Temporal and spatial variations in navigation data caused by deterministic and random phenomena can severely degrade integrated system accuracy unless suitable compensation is applied. An estimation procedure capable of modeling the statistical behavior of measurement and system error dynamics during real time operation has the capability of compensating measurements corrupted by stochastic effects. Several candidate estimators are investigated within a marine integrated system environment. A Kalman estimator is shown to provide the strongest benefits. A cautionary note is included concerning the effect modeling errors and numerical effects may have on estimation accuracy.