Improved discrete-time controller performance using the "control-dimension multiplier effect" of discrete/continuous control theory

The closed-loop performance of linear systems, using discrete-time controllers of the zoh-type, is inherently limited by the rank of the control-distribution matrix B/spl tilde/=/spl int//sub kT//sup (k+1)T/ /spl Phi/ ((k+1)T, /spl tau/)B(/spl tau/)d/spl tau/. It is shown that such performance limitations can be overcome, or significantly mitigated, without increasing the dimension of the control-vector, by using a new generalization of conventional discrete-time control called "discrete/continuous" control. Several examples are presented to illustrate the results.