Integrating geometallurgy in the simulation of processing plant responses and applying policy gradient for shovel allocation in operating mining complexes

Abstract

ABSTRACT Several short-term planning activities of mining complexes must be addressed simultaneously, such as extraction sequence, equipment allocation, digline definition, and destination decisions. Regarding stochastic mine planning, geological attributes and equipment performance uncertainties are more commonly addressed. However, processing response attributes at the ore treatment facilities also have uncertain components. This paper proposes a high-order simulation of processing plant responses given the blended ore attributes. First, it is included in the grade-control digline approach and later applied to a policy gradient-based framework that allocates shovels. A case study at a gold mining complex illustrates the benefits of the proposed approach.