An Adaptive Estimator for Time Varying Processes with Maneuvers.
Abstract
An adaptive discrete Kalman filter is developed which is based on an improved covariance matching technique. The filter quickly finds the unknown process noise covariance implied by the observed data. The process noise covariance is assumed constant. all other filter parameters are assumed known, although possibly time varying. This filter, used in conjunction with a multiple model filter, can be used to iteratively estimate noise statistics and detect maneuvers. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Aug 01, 1986
- Accession Number
- ADA172837
Entities
People
- Michael A. Schiefer
Organizations
- Air Force Institute of Technology