Optimum control of the Leeudoorn semi-autogenous milling circuit

intelligent control techniques, such as fuzzy logic, have been successfully applied to the control of a semi-autogenous (SAG) grinding circuit at the Leeudoorn Gold Mine. Benchmarking showed that circuit performance was enhan ced by the control system. Benefits included a significantly more stable ci rcuit, improved power efficiency, a reduction in the amount of steel grindi ng media required, as well as more consistent throughput. A software system , Grind-ACE, based on the intelligent autonomous agent-blackboard paradigm, was developed as an implementation platform. This approach has proved succ essful and flexible, allowing both intelligent and traditional control tech niques to be integrated. The control system has been implemented in a uniqu e manner with both stabilizing and optimizing capabilities, which has resul ted in a flexible and adaptable system.