Relative Response Array: A New Tool for Control Configuration Selection

Abstract

This paper is an attempt to overcome the limitations associated with a dynamic measure of process interaction the " Relative Response Array (RRA) " . The RRA was originally proposed for 2 × 2 plant models only. Through this paper we are proposing the four different versions of the RRA to make it a more generalized measure of closed-loop interaction. They are defined based on open and closed-loop step response of the plant model elements using controller-independent and controller-approach. To show the applicability of the proposed measures two different examples (a 2 × 2 non-physical benchmark problem and a 3 × 3 distillation column control problem) from refereed literature have been considered. The proposed measure successfully identifies the best control configuration in both the cases whereas the well known measure of process interaction the RGA fails in one of the cases. All the simulations are performed in Matlab/Simulink 7.