A Neural Network Controller Of A Flotation Process

The dynamic control of a froth flotation section is simulated through a neural network feedback controller, trained in order to stabilize the concentrate metal grade and recovery by applying random step changes to the feed metal grade. The results of the application example show that this controller seems to be sufficiently robust and a good alternative to handle a non-linear process.