Robust Control System Design using Direct Frequency Response

This paper describes a novel control system design methodology in which a process frequency response estimate, along with an associated modelling error, are used to design a robust parametric controller based on a user-specified, closed-loop transfer function. The controller parameters are obtained from an iterative procedure which minimizes the sum of squared differences between the estimated and desired open-loop frequency responses. The frequency response estimate of the process is obtained from an identification experiment using a frequency sampling filter model in conjunction with a batch least squares estimator. This identification technique produces a consistent estimate and an additive modelling error bound which is probabilistic in nature. Experimental results using a pilot plant heat exchanger are presented to illustrate the controller design.