A Hybrid NSGA-III and PSO Framework for Multi-Objective Ore Blending Optimization in Open-Pit Polymetallic Mines

Open-pit polymetallic mines commonly encounter challenges such as the asynchrony between mining and processing operations, large fluctuations in ore supply structure, and high haulage volumes, which lead to increased transportation costs and instability in processing plant feed grades. To address these issues, this study, driven by practical production requirements, proposes a two-stage hybrid optimization strategy that combines the global search capability of NSGA-III with the local intensification of particle swarm optimization (PSO), aiming to achieve the coordinated optimization of transportation cost minimization and plant feed grade maximization under constraints imposed by ore supply boundaries and processing plant capacity. To further identify the most suitable solution from the resulting Pareto-optimal set, the VIKOR multi-criteria decision-making method is employed to evaluate and select a blending scheme with optimal balance under the dual objectives of cost and grade. The effectiveness of the proposed approach is validated using a real-world production case, with experimental results showing that the optimized blending scheme achieves a cost reduction of more than 9%, while the gold grades of oxide and sulfide ores are increased to 2.40–3.16 g/t and 2.14–2.17 g/t, respectively, leading to a significant improvement in the overall plant feed grade. Compared with the actual weekly blending plan used in practice, the proposed method enables a comprehensive optimization of transportation cost, feed grade, and ore supply structure within a unified framework.