A heuristic approach to the design of linear multivariable interacting control systems

Abstract

The proposed approach aims to determine an input-output pairing configuration and controller settings guaranteeing a practically satisfactory disturbance attenuation for linear, interacting multivariable control systems. The argument is based on (a) a theorem which rejects input-output pairing configurations leading to structurally monotonic unstable systems with all controllers integrating, and (b) two heuristics which provide a choice between the remaining configurations and determine the controller settings. The heuristics are based on the concepts of the P-controlled multivariable systems critical gain and frequency. The paper is illustrated by two examples of two-variable control and one example of three-variable control. The proposed approach is well-suited for on-line tuning of interacting multivariable control systems. It was found to have significant advantages over conventional trial and error methods.