Tracking-filter structures for automatic track-while-scan surveillance systems / by A.W. Bridgewater. : Co24-3/7-1341-1981E-PDF

"In an automatic track-while-scan air-surveillance radar system, the role of the target-tracking filter is to support the plot-to-track association process by providing reliable estimates of the current track state, on which to base predictions of subsequent trackstates. The Kalman filter is the most general solution of the recursive linear mean-square estimation problem, its drawback being its computational cost An analysis of various one-dimensional forms of filter, derived from the Kalman, results in the complete specification of the a-f3(--y) and reduced-Kalman classes of tracking filter, which often form the basis for the design of the practical two- and three-dimensional trackers required inair-surveillance systems and which combine good track-following ability, ease of adaptation to changes in tracking conditions and low computational cost. Both recursive and steady-state adaptive versions of these one-dimensional structures are described. A functional relationship between the set of gain coefficients and a parameter which quantifies the current tracking conditions in terms of target manoeuvre uncertainty, radar measurement error and track update interval makes possible rapid and effective filter adaptation"--Abstract, page 1.

Lien permanent pour cette publication :
publications.gc.ca/pub?id=9.894801&sl=1

• Format MARC XML
• Format MARC HTML

Demander des formats alternatifs