Automatic design of digital controllers

Abstract

In this paper a procedure for the automatic design of digital SISO controllers is presented. The design is based solely on input-output data from the process, and the user does not necessarily have to make any design decisions. The first step of the design procedure is the identification of a process model and the estimation of its uncertainty. In the second step a controller is designed based on this model. The controller is fine tuned in such a way that robust stability is guaranteed when the estimated uncertainty from the first step is taken into account. The software for both steps is written in Matlab. The paper first describes the basic principles and tools of the design procedure. Then the software implementation is described. The last two sections contain examples and conclusions.<>