Аңдатпа
The issues of dilution reduction and marketable ore recovery in Sublevel Caving (SLC) are considered, using the Almaz-Zhemchuzhina deposit as a case study. An integrated method is proposed, combining sequential-stage ore drawing and digital blasting control via electronic initiation and adaptive parameter adjustment. The novelty lies in implementing a «monitoring – modeling – control» closed-loop cycle, utilizing Discrete Element Method (DEM) and Machine Learning (ML) algorithms to predict waste rock entrainment and optimize the drawing sequence. Field tests and calculations confirm a dilution reduction from 11.5% to 9.7%, decreased losses at the drawpoint floor, and an additional recovery of up to 1600 tonnes of ore. The methodology is practically applicable under conditions of complex orebody morphology and is recommended for analogous deposits