City Research Online

Abstract

A travelling-load, electrically heated furnace interfaced to a process control computer has been successfully controlled using an on-line recursive identification technique in a model-reference, sub-optimal control scheme. The problem of interaction among control loops has been largely mitigated by repeatedly updating the model parameters with every sampling instant, thus yielding a 'dynamic' model especially desirable during start-up operations of processes. It is demonstrated that modest computing requirements are required to implement the identification and control calculations through the use of the 'Sub-optimal' controller where optimization is performed over individual sample intervals (rather than over the whole of process operation time). A comparison is made with the performance of a conventional Proportional-Integral-Derivative controller from the points of view of ease of implementation, flexibility and control action obtained. The performance of both schemes are seen to be satisfactory.

Computer programs have specially been written in the Argus 500 language ASTRAL to control the furnace in real-time. Operator/machine interaction is via a relatively simple conversational-mode program. A modular programming approach has been attempted wherever possible. Pully documented listings are included in the appendices.

A survey of several relevant control techniques and industrial furnace users is included in order to place the implemented techniques in perspective and point to future lines of research.

Publication Type:	Thesis (Doctoral)
Subjects:	Q Science > QA Mathematics > QA75 Electronic computers. Computer science
Departments:	School of Science & Technology > Department of Computer Science School of Science & Technology > School of Science & Technology Doctoral Theses Doctoral Theses

Export

Downloads