BIL98/59 - Neural networks and advanced methods for monitoring and control of flotation plants

From 21-12-1998 to 20-12-2001

Description

Although flotation processes are notoriously difficult to model from first principles, preliminary work at the Univ. of Stellenbosch has shown that neural networks can be used to considerable advantage in order to monitor an control plants, provided that process knowledge can be captured effectively. For example, by making use of machine learning techniques the features of the surface froths of flotation cells can be used to construct representations of the behaviour of a plant. These systems need to be implemented on an industrial scale. Moreover, the dynamic behaviour of these plants would require continuous adaptation of plant models. In the project, it is envisaged that data currently obtained on-line in the form of digitized images of the froth structures of an industrial flotation plant, as well as other relevant metallurgical parameters be used as a basis for system identification and the development of a decision-making systems for the operators.

Team

• Bart Motmans, Coordinator
• Joos Vandewalle, Promotor

Financing

Funding: Vlaamse overheid - Vlaamse overheid

Program/Grant Type: BIL - Bilateral Scientific Cooperation