| Column flotation is an important separation process in the mineral processing industry. Its performance is measured by the grade and recovery of the valuable mineral in the concentrate. However, the on-line optimization of the metallurgical variables remains a goal difficult to reach for the industry and therefore the control of intermediate variables is more realistic and necessary, nevertheless the unreliability of the on-line sensors does not make possible to have a good control. This work presents the instrumentation and such an advanced control strategy for a three-phase system pilot flotation column. The design and implementation of a multivariable predictive controller with constraints is described. The controlled variables are the froth depth and the gas holdup. A modified and improved algorithm of a conductivity profile measures the froth depth. The gas holdup is measured and compared either by hydrostatic pressure or by conductivity with a new method of suction. A controller tuning method based on the scaling of the process model and a controller modification for integrating processes are also described. |
