Factorization Methods for Discrete Sequential Estimation

By Gerald J. Bierman

Book Overview

This estimation reference text thoroughly describes matrix factorization methods successfully employed by numerical analysts, familiarizing readers with the techniques that lead to efficient,... This description may be from another edition of this product.

• Format: Paperback
• Language: English
• ISBN: 0486449815
• ISBN13: 9780486449814
• Release Date: May 2006
• Publisher: Dover Publications
• Length: 258 Pages
• Weight: 0.11 lbs.
• Dimensions: 0.5" x 5.5" x 8.5"

Customer Reviews

Classic Square Root Filter Reference

A square root filter is a must in recursive estimators having large numbers of states or measurements, or in applications where numerical problems are apparent. This monograph presents three square root Kalman filter implementations -- the square root covariance filter, the UDUT filter, and the square root information filter (SRIF). The ordinary Kalman filter and Joseph stabilized form are also included for comparison. Although the SRIF is presented convincingly here as the best choice in most applications, the effect of this book has been to make the UDUT commonplace as the path of least resistance. The book includes tutorials, derivations, equations for operation counts, and pseudocode algorithms. A well-considered square root filter provides a solid foundation for the most complex multiple hypothesis tracker (MHT), sensor fusion algorithm, and is a must for an INS or autopilot, because it provides a rock-solid base estimator. This avoids common misapplications of development focus such as unnecessarily large plant noise and distractions to fix a less robust Kalman filter. This reference is invaluable in constructing and maintaining square root filters, or in converting to square root filters.