Adaptive modeling of mining schedule using genetic algorithm in a dynamic environment

The operation of large mining clusters faces significant challenges, as long-term planning is often based on constant environmental conditions, ignoring uncertainties in the economic environment, geological uncertainty, and the mutual influence of enterprises. However, the current realities of the global market and the challenges of the industrial crisis require consideration of dynamic conditions, such as price and demand fluctuations, during mining operations. This study proposes a method for multi-factor optimization of mining operations based on a genetic algorithm. The methodology involves the mathematical formalization of a mining cluster system combining open pit mines and industrial deposits, using an evolutionary approach to solving nonlinear optimization problems that take into account ore quality and mining costs. The study identified optimal parameters for the optimization algorithm – specifically, a crossover value of 0.7–0.8 and a mutation probability of 5% – to ensure the accuracy of design decisions and prevent optimization from drifting toward local minima. The proposed model determines annual production blocks, storage volumes, and processing of man-made deposits, enabling dynamic adjustments to cutoff ore grade values to minimize overall costs. Unlike traditional deterministic approaches, the proposed methodology provides flexibility in planning, enabling long-term reductions in production and storage costs while maintaining stable ore quality. The results obtained can be used by engineering design organizations for adaptive, dynamic long-term design of mining facilities and regional clusters, as well as for improving the specialized software used for this purpose.