The impact of filtering on maritime and flight operations
Abstract
An analysis of the computational techniques involved in automated filtering of radionavigational data is presented. The Kalman filter offers simple, recursive data processing to minimize errors. Use of Kalman filter has made the application of Bayesian estimators, maximum likelihood, and least squares methods of doubtful value. Linearity is a necessary condition for Kalman filtering, and can be disrupted by the absence of an exact definition of the covariance matrix, by a divergence in the calculations, by an incorrect initialization of the data, by truncation errors, by scaling errors, or by resynchronization of the data input with respect to the signal processing speed. The necessity of building repeatability into equipment is stressed, and a sample problem is worked out in terms of the geometry of localization between signal stations.
 Publication:

Navigation Paris
 Pub Date:
 January 1983
 Bibcode:
 1983NavPa..31...39P
 Keywords:

 Kalman Filters;
 Radio Navigation;
 Air Navigation;
 Analytic Geometry;
 Covariance;
 Signal Processing;
 Surface Navigation;
 Truncation Errors;
 Space Communications, Spacecraft Communications, Command and Tracking