Flotation Velocity and Grade Control Using Cascaded Model Predictive Controllers

Abstract

Flotation of valuable minerals remains a key unit operation on the minerals processing industries. In a modern automated plant, the control handles are in place to stabilize and optimize the flotation process. The challenge has been the reliable measurement of the required control parameters. This paper describes the synthesis of image processing, online analysis and cascaded model predictive control (MPC) to optimize a copper roughing circuit at the Kansanshi operation in Solwezi, Zambia. Implementation challenges are discussed, and key results highlighted.