Developments in adaptive control and its potential in the minerals industry

Much of the process control effort in the mineral industry uses proportional-integral-derivative controllers that require periodic retuning. Retuning hundreds of controllers typically employed in a single plant is a formidable task for the control engineer. Adaptive controllers offer an alternative to this problem, since they are designed to tune themselves. Four schemes of adaptive control-gain scheduling, model reference adaptive control, self-tuning regulator, and optimal adaptive controllers-are discussed with process examples. The control of a ball mill circuit with a multivariable self-tuning regulator is illustrated. This controller encompasses the advantages of the other three controllers and is applicable for a wide range of process variability.