The MLE Tracking Algorithm: A Summary and Error Analysis.
Abstract
This report discusses, in general, least-squares tracking algorithms and in particular the maximum likelihood estimator (MLE). Its primary concern is a detailed error analysis both from a deterministic point of view (sensitivity study) and a stochastic viewpoint (state variable covariance matrix). The latter results are shown to be equivalent to a Cramer-Rao bound, and their relationship to Kalman filters is cited. Various subtleties of interpretation are discussed including several theorems on confidence ellipsoids. Examples generated by computer software based on the theory are also presented. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- May 15, 1981
- Accession Number
- ADA103427
Entities
People
- M. J. Shensa