DocumentCode
813772
Title
Choice of coordinates and computational difficulty
Author
Johnson, George W.
Author_Institution
IBM Federal Systems Division, Burlington, MA, USA
Volume
19
Issue
1
fYear
1974
fDate
2/1/1974 12:00:00 AM
Firstpage
77
Lastpage
78
Abstract
Computational difficulty can be significantly reduced when optimal control and estimation problems are formulated in well-chosen coordinate systems. It is shown that every LQG (linear-quadratic-Gaussian) problem can be formulated in coordinates such that principal parts of the solution are either well-posed or ill-posed for numerical solution.
Keywords
Estimation; Optimal control; Computational modeling; Digital filters; Distortion measurement; Eigenvalues and eigenfunctions; Ellipsoids; Length measurement; Observability; Phased arrays; Radar tracking; Rotation measurement;
fLanguage
English
Journal_Title
Automatic Control, IEEE Transactions on
Publisher
ieee
ISSN
0018-9286
Type
jour
DOI
10.1109/TAC.1974.1100471
Filename
1100471
Link To Document