A New Procedure to Account for Performance Interaction in Multivariable Systems

Some new results in linear multivariable systems interaction analysis are reported in this paper. Using a newly proposed performance interaction measure which can be applied to the transient state as well as to the steady state, a systematic procedure to reduce interaction is presented which efffectively improves closed-loop performance and at the same time adjusts system stability robustness. The analysis and design procedure is based on the matrix singular value decomposition and the polar decomposition of a transfer-function matrix.