Exact multivariable control design using SISO methods: Recent results

In this paper we describe some very recent methods developed by the authors to design multivariable controllers exploiting single input single output (SISO) design methods. These methods allow the designer to accommodate several useful design criteria such as low order, steady state tracking and disturbance rejection, overshoot and settling time control and gain and phase margin, The main tool is the Smith-McMillan form of the transfer function which can be used to reduce an n×n multivariable design problem into an equivalent set of n single input single output problems. An additional advantage of the method is that a state space realization is not needed and many criteria that are beyond the reach of quadratic optimization can be achieved.