关键词: 露天矿　, 精细化配矿　, 配矿优化模型　, 求解方法　, 配矿系统　, 不足　, 展望

Abstract: In response to the challenges such as the high difficulty in the comprehensive utilization of low-grade resources and significant fluctuations in ore grade during open-pit mine production,focusing on the key difficulties of refined ore blend ing,and combined with the research accumulation of the research group in the field of ore blending optimization,the evolution process and characteristics of ore blending optimization models from single-objective optimization (including production effi ciency,grade deviation,and ore quantity) to multi-objective collaborative optimization have been sorted out.The applicable scenarios,advantages and disadvantages of three mainstream solution methods,namely classical mathematical programming,intelligent optimization algorithms,and fuzzy mathematics,have been summarized.The application practices of mainstream do mestic and foreign mining software (e.g.,Minesched,DIMINE,3DMine) as well as the development achievements of intelli gent ore blending systems in terms of technical architecture and full-process functional modules have been discussed.The re sults show that ore blending optimization has shifted from single-objective optimization to multi-objective collaborative optimiza tion,intelligent optimization algorithms have become the core tools for addressing complex constraint scenarios,and mining soft ware and intelligent ore blending systems have promoted the transformation of ore blending optimization from theoretical re search to large-scale and engineering applications,thereby improving the efficiency of ore blending decision-making.On this basis,the shortcomings of current research have been analyzed,and the development directions for the "15th Five-Year Plan" period and even a longer time frame have been prospected.It is identified that current limitations primarily stem from the un clear mechanisms of grade reconstruction induced by blasting operations,the insufficient refined consideration in formulating ore blending plans,and the lack of dynamic adjustment capabilities within ore blending systems.Specifically,these deficiencies manifest as:① The coupled effects of uncertainties in geological estimation and blasting operations are difficult to accurately quantify;② The characterization of blasting pile homogenization and the scale of local mining operations are difficult to be ef fectively matched;③ The intelligent identification of the temporal sequence dependency of the mining unit still needs to be solved;④ The mathematical quantification of on-site process criteria and its integration with the ore blending optimization mod el require deeper integration;⑤ The dynamic emergency correction mechanism for the ore blending plan is yet to be perfected; ⑥ The robustness of the ore blending system under special conditions is difficult to be effectively guaranteed.Future research should focus on three core directions:re-creation of blast grade models and detailed division of ore blending units,identification of the interdependencies of ore blending units and refined modeling of ore blending plans,dynamic adaptive decision-making for ore blending and multi-modal perception of extreme working conditions.Specifically,it covers six aspects:① Grade model reconstruction based on the coupling of blast displacement monitoring data and geostatistics;② Refined delineation of ore blending units based on geological and mining constraints;③ Automatic identification of mining dependencies among ore blending units;④ Optimization modeling of ore blending plans embedding spatial topological and mining safety constraints; ⑤ Dynamic intelligent decision-making for ore blending based on technologies such as reinforcement learning;⑥ Development of multi-modal fusion hardware systems for sensing and positioning in ore blending under extreme working conditions.Through these technical breakthroughs,systematic support and practical references can be provided for the improvement of the ore blending optimization theoretical system and the efficient implementation of engineering practices.

Key words: open-pit mine,refined ore blending,ore blending optimization model,solution method,ore blending system, limitations,prospects