Adaptive Estimation Algorithms.
Abstract
The study considers the problem of estimating the states of a linear discrete dynamical system when the covariance matrix, R, of the stationary white sequence corrupting the measurement and/or the covariance matrix, Q, of the stationary white input sequence are unknown. Two new adaptive estimators, called the Reprocessing Filter (RF) and the Maximum A Posteriori (MAP) estimator, are developed which jointly estimate the state variables and the unknown R and/or Q. The new feature common to both estimators is the use of easily implementable estimators of R and/or Q in a reprocessing configuration with the Kalman-filter algorithm. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Dec 01, 1970
- Accession Number
- AD0720394
Entities
People
- Larry J. Levy
Organizations
- Iowa State University