Design of a Wear Estimator for liners in SAG Mills using ANFIS modeling

In grinding, the SAG mills are used because of its capacity and short retention time for the processing of minerals and metals. SAG mills contain liners inside its structure. The liners crush the mineral and must be replaced each 6 months because they deteriorate during the crushing process. The measure of liners wears in SAG Mills requires a shut down operation, that implies high costs when it is necessary to change liners suddenly. This paper proposes the use of an Adaptive Neuro Fuzzy Inference System method (ANFIS) for modeling a wear estimator for liner in SAG mills, which will train the network to obtain adequate indicators for the comparison in different configurations. The best response is obtained with a network using accumulated input values of variables taken from the SCADA system.