Noninteracting control by output feedback and dynamic compensation

The design of noninteracting multivariable control systems by slate feedback has received considerable attention recently, and an extensive theory is now available. One area, however, that requires further clarification concerns the case where the state vector is not accessible to measurement, and feedback of output is inadequate for the purpose of decoupling. In this paper a new technique is described by means of which it is possible to decouple a system which is not olherwise decouplable by proportional feedback of the output vector. The technique is based on the addition of a dynamic controller in the forward path, and is analogous to cascade compensation in single-loop systems. It is shown that a necessary and sufficient condition for the success of this method of decoupling control is that the Original system should satisfy the condition for state feedback decoupling.