Application of Fuzzy Cognitive Maps and Run-to-Run Control to a Decision Support System for Global Set-Point Determination

The final output of modern processes is significantly influenced by the selection of the set points of the process variables, as they fundamentally impact the product quality characteristics and the process performance metrics. This paper proposes a decision support system structure based on models built using fuzzy cognitive maps, an optimization problem that employs these models as constraints, and an observer that augments the optimization problem and allows to implement a run-to-run control approach, enabling the system to correct the suggested set-point values in case of a mismatch between the desired process output and the actual one. An application example related to an industrial process is included to illustrate the proposed approach.

I. Introduction

The general objective of a production process can be regarded as the maximization of the revenue of the activity. In order to address this objective, a series of concerns naturally arises: first, the characteristics of the good to be produced must be defined; second, the values of the process parameters that allow to achieve these characteristics must be determined, and, finally, the process must be run assuring that these parameters and, more importantly, the final product characteristics, remain at their prescribed values.