The Software Architecture of a Decision Support System for Process Plant Instrumentation

Abstract

In this work we present the framework of a new software package to carry out the classification of the state variables in a process plant in order to determine the most convenient instrumentation configuration on the basis of adequate mathematical models of plant behaviour at steady state. We have designed and developed robust and efficient techniques for each step of the categorization task, the key modules being those for observability and redundancy analysis. The individual performance of each module with both academic and industrial examples was successful, revealing comparable or even better results than other existing techniques. Key aspects, such as modularity, user-friendliness, reliability, flexibility, expansibility, functional scalability, and stand-alone capabilities, were taken into account for the design of the integrated software.