Underground mine planning: stope layout optimisation under grade uncertainty using genetic algorithms

Abstract

Stope layout optimisation finds a technically producible orebody portion that maximises the profit of the mining operation based on the stoping method used. A three-stage stochastic optimisation model combining genetic algorithms (GA) is proposed to account for grade uncertainty. The first stage computes the stope layout uncertainty, the second stage creates average design and their feasibility evaluation breeds the initial population, and the third stage uses GAs to improve this initial population over generations. The approach generates higher profit, less planned dilution, and a robust stope layout that is insensitive to orebody grade fluctuations.