PREDICTIVE CONTROL OF ALUMINUM ELECTROLYTIC CELLS USING NEURAL NETWORKS

In this work, neural networks are built and trained to be used in a pr edictive control scheme for the aluminum electrolytic cell. An efficie nt control of the cell requires the knowledge of predicted future valu es of the decision variables in order to enable the standard (nonpredi ctive) control logic to take anticipated actions to prevent the anode effect, a destabilizing event occurring during cell operation. The net works are first trained on data obtained from a computer simulator of the cell prior to undergoing further on-line learning. Trained to pred ict the cell resistance and the resistance's trend indicators, the net works are applied to the control of cells in different cell states, wi th a view to preventing anode effects, the latter being deliberately i nduced by reducing the alumina feed rate or reducing the feeding frequ ency and duration. Results show that, with neural-predictive control, anode effects can be avoided, which should result in increased thermal stability, decreased power consumption, and reduced fluoride emission s. Further, the on-line learning capacity of the networks offers a goo d perspective for their application to other complex industrial proces ses as well.