Advanced Process Control of a Flotation Column

Flotation cells a solid-solid separation process based on the physical and chemical
properties of the mineral particle surface. The process is widely used in the mineral
processing industry for a low grade and finely disseminated ores to recover the
valuable mineral. It is also appliedin recycling and solvent extraction process. These
are several problems associated withthis process namely:
a) Poor product recovery due to inconsistent froth stability
b) Lowproduct grade attributed to recover of undesired hydrophilic particles
c) High energy andmaintenance costof mechanical agitator
d) Poor control of the cell's level.
The main objective of this research project is to study a possible control type to
improve the level control mainly in advance control process. It definitely will
enhance the recovery and purity of precious mineral from the ores. There were 5
control types being evaluated in this study which are feedback, feed-forward,
cascade, smith predictor and fuzzy logic control.
The research work began with developing model of the flotation column process
using simulink toolkit within MATLAB software. It is succeeded with development
of the abovementioned controlleronto the process model. For those control that uses
conventional PID algorithm, similar tuning constant were applied. Each of the
control types were subjected to set point change on the level and disturbance (.i.e.
ratio of valuable and waste within the feed). The performance for every control type
was evaluated and trends were compared.
In conclusion, cascade control provides best performance both performance in set
pointchange and rejecting disturbance.