Design  | SFC  |  NOS  |  Code |  Diary  | WWW |  Essay |  Cisco | [Home]

 

  

[Bills Home]
[Old page]

General

Abstracts
Group Objectives
PhD/MPhil
Apply for a PhD

Agents

Agent-based, distributed security system
Mobile Agents in Network Applications
Agent-based Fault Diagnostics

Resuse in Dis Sys

Generic framework

Reuse in embedded systems

Encryption

RSA Method

General Networking

Gigabit Ethernet

Programmable Router

 

 

 

Integration of Intelligent Methods in the Design of Process Control
Researcher:

 
Date completed:

 
Funding:

 
Higher-degree:

 
Papers published:

 
Presentations:

 

Modern large-scale industrial processes are becoming more complex, due to new technology and the high-quality demanded from clients. Industry must continually monitor and change to compete successfully in a free market. New technology is associated with increased complexity, of design, understanding, use and management. Industry must continually invest in this equipment and its users, to prevent falling behind competition.

New technology has accelerated the evolution of intelligent systems such as Neural Networks, Fuzzy Logic and Expert Systems. These intelligent systems in theory may enhance or replace conventional systems such as Three-Term Controllers, Ratio Controllers, Cascade Controllers or Sequential Logic. However, the use of such systems in large-scale process systems is still extremely rare. Academia and industry have looked at using Neural Networks, Expert Systems and Fuzzy Logic to resolve process control and alarm handling problems [5-7]. However, these have been shown to be successful in only a few isolated industrial process units, as they are either expensive and/or time consuming to implement. These may also require separate systems to operate, which must interfaced to existing hardware systems. The success of such systems is difficult to quantify and therefore makes industry reluctant to invest in them. To enable the next stage of evolution of intelligent systems, they require integration with standard process control systems and comprehensive on-site testing, such as:

  1. Integration. This requires mechanisms to highlight, which points are to be linked into intelligent systems. These mechanisms will require in-depth understanding of the process. During the conceptual design, many studies are normally carried out. These may include, Hazard and Operability (HAZOP) studies, Hazard Analysis (HAZAN) studies, Risk Assessments, and so on [1-4]. Bringing this information together with the design criteria for intelligent systems will provide a tool and a knowledge base, which can be utilised to highlight link points. Link points are the interface points between the process and the intelligent systems.
  2. On-site testing. Intelligent systems require the system to be flexible and comprehensible. Flexibility enables an intelligent system to be monitored, analysed and optimised. Any system used in a large-scale processes requires this flexibility to absorb the unique characteristics of plant and equipment. When installing a new plant, there is little time to rectify obvious errors. It thus essential for this to be done as efficiently as possible and allowing personnel on varying levels of awareness to confidently use the intelligent system. To achieve this novel techniques in representation of the intelligent system will be developed.

The main aim of the research is to investigate the requirements for integrating intelligent systems into the synthesis and operation of large-scale industrial process control systems. This will involve developing tools for identifying link points, which integrate intelligent systems with conventional systems, and to examine different methods of representing intelligent systems, their interfaces and hierarchy.

References

  1. Part 15 of the Institute of Petroleum Model Code of Safe Practice in the Petroleum Industry. Area Classification Code for Petroleum Installations. John Wiley & Sons. March 1990.
  2. Project Procedures Manual. Safety and Operability Review Procedure. EniChem Elastomers UK Ltd. May 1989.
  3. Hazards Forum. Safety Related Systems Guidance for Engineers. The Hazards Forum. March 1995.
  4. Quantified Risk Assessment: Its input to decision making. Health and Safety Executive. 1994.
  5. Artificial Neural Networks for Intelligent Manufacturing. Cihan H. Dagli. Chapman & Hall. 1994.
  6. Introduction to Expert Systems. Peter Jackson. Addison-Wesley Publishers Ltd. 1990.
  7. Logical Foundations of Artificial Intelligence. Michael R. Genesereth. Nils J Nilsson. Morgan Kaufmann Publishers Inc., 1987.