Using Fuzzy Relational Models for Control

In this paper the use of Fuzzy Relational Models (FRMs) for model-based process control is investigated. The work presented here is centred around the development of a Fuzzy Internal Model Controller (FIMC) for a simulated multi-variable pH process. The paper begins with a brief introduction which sets the scene for the work. There then follows a description of the pH process and the particular problems it poses for control. The identification of a reasonable process model is investigated and the difference between a good model of the process and a good model for control is highlighted. The FRM is then fitted into a standard IMC controller structure and experimental results are presented demonstrating good control.