Skip to Main Content
Optimizing the final grinding production indices (GPIs), which include the product particle size and the grinding production rate, to meet the overall manufacturing performance requirements is the main function of automatic control of a grinding circuit (GC). However, the complex and time-varying nature of the GC process dictates that these GPIs cannot be optimized solely by the lower-level distributed control systems (DCS), therefore an operator is often incorporated to manually determine the set-points for the DCS using his/her operational experience. With a human being involved, the performance and even the safety and stability of the GC operation is subject to human errors. Focusing on this practical challenge, this paper proposes an intelligence-based supervisory control strategy that consists of a control loop set-point optimization module, an artificial neural network-based soft-sensor module, a fuzzy logic-based dynamic adjustor, and an expert-based overload diagnosis and adjustment module to perform the control tasks for the GC system. This hybrid system can automatically adjust the set-points for the DCS-controlled grinding system in response to the changes in boundary conditions or the imminent overload conditions, thereby eliminating the need for an operator. Practical applications have shown the validity and effectiveness of the proposed approach.