A dual-rate self-tuning pole-placement controller

Abstract

A dual-rate self-tuning pole-placement controller is proposed. It allows the use of a larger sampling interval for on-line parameter estimation and a smaller sampling interval for control. The conversion of model parameters from one sampling interval to another can be readily performed using a pole-zero mapping technique. The major problem encountered is that the process zeros at the slow sampling rate cannot be estimated accurately by using input-output data obtained at the fast sampling rate. The proposed solution of fixing the zeros at z=-1 while keeping the DC gain the same is supported by analysis of the Nyquist plot. The main merits of the dual-rate self-tuning control system are in the simultaneous achievement of robust estimation and reduced computation load and in improved performance in the regulation of deterministic and stochastic load disturbances. Its performance is substantiated by digital simulation and experiments on a pilot plant. >