Design Of A Knowledge-Based Intelligent Supervisory Control

The metals industry and plant equipment suppliers are constantly striving to obtain a competitive edge. Level I dynamic control products from British industry are world-class, but it is recognised that the level II supervisory control systems can be improved. This project seeks to achieve this by adopting new intelligent technologies to utilise plant knowledge. Expert systems have been selected as an appropriate technology as they capture knowledge in a robust and industrially proven fashion. The design of the supervisory controller will focus on this knowledge handling strength of the Expert System, by using it to augment an existing model based adaptive control system. To enable improvements in the control of plant the design expands the scope of plant control, focusing on important issues and uses the commodity of knowledge. The project seeks to prove the benefits of utilising the Expert System technique in industrial supervisory control, by considering the quantitative effect of employing an Expert System in a real life situation. This paper will give background to the project, motivate the use of the Expert System and describe its design