PAPYRUS extends diagnostics from a purely passive to an active approach controlling the entire plant.
The first objective of the project is plant-wide application of asset management. It lifts the focus from single-component diagnostics to a plant-wide view of both asset performance and asset condition. Secondly, because diagnostics need to result in actions, diagnostic information (asset information) is used to generate control actions. Control refers to all levels of the automation pyramid, i.e. production control, plant maintenance, as well as field level control. A third and necessary objective is to achieve and maintain this in a modular and efficient way.
The starting point for plant-wide diagnosis is the collection, harmonization and aggregation of the available component diagnoses in order to be able to realize a plant-wide approach. One essential aspect in this work is the integration of both condition information and performance assessment.
A decision support system will also be developed that maps plant information, together with constraints and requirements, in order to generate and propose plant actions that satisfy production and other requirements. Implementation of these concepts can only be done using modern software architecture in accordance with available standards (modeling interfaces). One important basic principle will be the ease of use. Large-scale applications require simple, standardized solutions for sustainable operation.
A modern production plant is a highly complex and interconnected entity. The concepts developed in the project will, in a straightforward manner, be applicable to other industries, and also to the site level once the plant level solution is in place.
The PAPYRUS project will contribute to both the development of the FDI/FTC (Fault Detection and Isolation/Fault tolerant Control) theoretical and methodological background for the large scale systems as well as prototype design for an important novel process industry application.
It must be noted that the architecture described here extends current control system applications quite far since it not only involves traditional diagnosis and control but also plant performance monitoring and service activities. It hence leaves the area of traditional component-based activities and takes a real plant-level view. This view sets the basis for optimal economical benefits.
Timely the project will be executed in four main periods, with well-defined milestones at the end of each period:
Period 1: Collection of requirements and detailed specification of research and development work.
Period 2: Development of methods for structural analysis, functional modelling for the system and FDI/FTC in the networked system environment.
Period 3: Integration for the embedded PAM software platform.
Period 4: Application to an end user infrastructure and plant.