Grinding mill modeling and control: Past, present and future

This paper addresses issues relating to the modeling and control of grinding mill circuits as used in minerals processing, with a particular emphasis on (semi) autogenous grinding or SAG mills. SAG mill circuits are generally difficult to control due to the presence of strong external disturbances, poor process models and the unavailability of important process variable measurements. In addition, it is difficult to independently control important variables such as the product particle size, throughput and product density, because independent control of the amount, size and hardness of the grinding medium in the mill is not possible. Hardness and size changes therefore introduce significant disturbances that necessitate the use of feedback control in order for the plant to operate efficiently. Technical and economic objectives of milling circuits will be discussed and formulated. The issue of modeling the milling circuit for control purposes will be presented and a recently developed model highlighted. The application of μ-synthesis and robust nonlinear model predictive control to a grinding mill circuit will then be discussed. Milling circuit controllers generally require tools peripheral to the controller itself to function optimally and such tools are briefly discussed. The paper concludes with some future research challenges.