Intelligent control of the feeding of aluminum electrolytic cells using neural networks

摘要

To be efficient, the control of alumina feeding of theelectrolytic cell must be based on cell resistance, aluminaconcentration, and cell state. Most control schemes now in use arebased on cell resistance only, and, thus, constitute an open-loopcontrol tat lacks robustness because their decision criteria are notexplicitly tied to concentration nor to cell state.