Disturbance rejection in autotuners: An assessment method and a rule proposal

Disturbance rejection is a primary objective in many industrial control loops, thus a relevant goal for autotuning controllers. Nonetheless, autotuning has invariantly to cope with a reduced amount of process information. As a consequence, with the standard single-loop structure typically adopted in the addressed context, effective disturbance rejection calls for strong feedback, and therefore the solutions available to date fall sometimes short of perfection. This paper discusses the matter basically from a methodological standpoint, evidencing some structural reasons for the observed shortcomings. The result is a synthesis approach improving rejection performance with respect to existing and well established tuning rules, on a rigorously sound basis. Simulation examples are presented to support the proposal.