DID YOU KNOW? DTIC has over 3.5 million final reports on DoD funded research, development, test, and evaluation activities available to our registered users. Click HERE
to register or log in.
THE DESIGN AND SYNTHESIS OF LINEAR MULTIVARIABLE SYSTEMS WITH APPLICATION TO STATE ESTIMATION.
STANFORD UNIV CA STANFORD ELECTRONICS LABS
The theme of this paper is the analysis, design, and synthesis of multivariable linear systems with a particular view toward applications in state estimation. A canonicalization procedure that provides a rationale for the design and analysis of constant linear multivariable systems is developed. When a multivariable system is excited by state and observation noise, a stochastic filter is usually required to estimate the system states. A technique is developed that permits simple comparisons of the performances of Kalman, Wiener, and general constant linear filters. Both transient and steady state error behavior are considered, and bounds that provide convenient guidelines for filter selection are obtained. Determination of a set of state variables for a constant linear multivariable system being simulated or synthesized can be made according to a minimum eigenvalue sensitivity criterion. A procedure that provides the synthesized system with a sensitivity canonical representation is established. Author