An Optically Implemented Kalman Filter Algorithm.
Abstract
An extended Kalman filter algorithm which incorporates an enhanced correlator/linear measurement model and a nonlinear target acceleration dynamics model is described for use in a pointing and tracking system for high-energy ground-based laser weaponry. The measurement model used in the filter combines the computational benefits of a correlation algorithm with the statistical accuracy available from a Kalman filter. The dynamics model, based on a constant turn-rate target acceleration model, is deemed to be a better representation of the true target dynamics than a linear first-order Gauss-Markov target acceleration model for the cases of interest. Optical processing techniques are completely specified for the correlation stage to perform the required correlation in real time, and the filter stage to perform the linear algebra required for the Kalman filter. This extensive use of optics allows the development of two tracking algorithms based on the same models: a FLIR-constrained tracker with a 30 Hz frame rate and an unconstrained tracker with a 100 Hz frame rate using real-time sensors in place of the FLIR. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Dec 01, 1983
- Accession Number
- ADA141126
Entities
People
- W. A. Roemer
Organizations
- Air Force Institute of Technology