Abstract: The conventional practice of relying on manual experience for ore blending in open−pit bauxite mines consistently leads to unstable feed grades, inefficient resource recovery, and increased downstream processing costs. To address these limitations and achieve precise operational control, this study developed an intelligent ore blending methodology tailored to complex production environments, with the explicit objective of minimizing ore grade fluctuations and enhancing overall planning efficiency.By established a multi−constraint optimization model aimed at minimizing fluctuations in the aluminum−silicon ratio,the model incorporates two critical operational constraints: a continuous mining constraint to reduce frequent transitions between mining faces and the resulting equipment downtime, and a balanced mining constraint designed to synchronize the extraction progress of each mining zone with the planned sequence and intensity. To address this complex nonlinear programming model, we employed two optimization methods: nonlinear programming (NLP), which focuses on precise local optimization, and the non−dominated sorting genetic algorithm III (NSGA−III), which searches for the global Pareto optimal solution set.The proposed system was tested using three years of actual production data from a large−scale open−pit bauxite mine in Guangxi. Experimental results demonstrated a significant improvement over manual methods. The traditional approach yielded an A/S fluctuation rate of 3.603%. In contrast, the NLP−optimized schedule reduced this rate to 0.315%, while the NSGA−III−derived solution further decreased it to 0.019%—an improvement of nearly two orders of magnitude. Beyond grade stabilization, the optimized plans also enhanced operational smoothness by reducing unplanned equipment idle time and improving the stability of mining operations across different mining areas.This study demonstrates that the proposed intelligent ore blending model and solution strategy are effective methods for stabilizing feed grade in complex mining operations. NSGA−Ⅲ is a practical ore blending optimization algorithm that helps mining planners balance equipment productivity and resource utilization while maintaining consistent grade. It holds significant engineering application value and shows promising prospects for widespread adoption.