INTEGRATED ROBUST IDENTIFICATION AND CONTROL OF LARGE-SCALE PROCESSES

We propose the use of pseudo-singular values, which are closely relate d to singular values but are allowed to have sign, as a convenient app roach for developing techniques for the identification and control of large-scale processes. Steady-state controllability can be assessed di rectly in terms of the pseudo-singular values. It is shown that to con trol an output disturbance direction with zero steady-state error it i s necessary to correctly identify the sign of the corresponding pseudo -singular value. At the identification stage, this motivates the estim ation of confidence intervals for the pseudo-singular values from inpu t-output data. A controller with integral action should not attempt to manipulate in process input directions corresponding to output distur bance directions that cannot be controlled with confidence. These prin ciples motivate a controller structure appropriate for providing the r obust control of poorly conditioned large-scale processes. Any control ler design technique can be applied to produce a controller with the p roposed controller structure. The controllability results and the inte grated identification/controller design procedure are illustrated usin g an industrial paper machine example.