Inverse-Nyquist-array design with minimum sensitivity

The inverse-Nyquist-array method for multivariable control-system design is extended to systems subject to parameter variations. Sensitivity of the inverse Nyquist array to plant-parameter variations is introduced and a worstcase design approach is adopted for achieving diagonal dominance. The method determines the best values of elements of a precompensator matrix which maintains the system diagonally dominant for parameters within a specified domain. The method is applied to the design of a precompensator for a multivariable a.c. turbogenerator model with five variable parameters.