Ore Blending Decision Of Polymetallic Open-pit Mine Based On Improved NSGA-? Algorithm

Mineral resource restriction and ecological civilization construction promote the transformation of traditional mining industry.The application of big data and artificial intelligence accelerates the traditional mining industry to move towards intelligent and automated mines.Efficient utilization of mineral resources has become an urgent demand of the mining industry.With the current situation of multi-component associated production in open-pit mines,the comprehensive recovery of multi-metal associated production accounts for a large proportion in the output value of enterprises.The complex geological conditions and large fluctuation of ore grade have significant effects on the utilization rate of resources,so rational ore blending as a means to improve the utilization rate of resources shows great significance.At present,the study of single ore type and single objective ore blending cannot meet the needs of mining enterprises.The polymetallic multi-objective ore blending model and its solution are very important for the ore blending optimization.Although the solution of ore blending model has been transformed to intelligent calculation,most of them are converted to single objective solution,ignoring the conflict between objectives.Based on this situation,this paper aims to study reasonable and efficient ore blending work to improve the comprehensive utilization rate of ore and increase the economic benefits of enterprises.Thus,the ore blending problem of polymetallic open-pit mine is studied as follows:(1)According to the development status of the mining industry,the significance of ore blending research is expounded.This paper reviews the domestic and foreign scholars' solutions to the ore blending problem,summarizes the research status and shortcomings of ore blending,and points out that it is an urgent need to improve the utilization of mineral resources.The ore blending principle is briefly summarized simultaneously.(2)The research on ore blending in the past 10 years is analyzed and summarized from two directions of ore blending model and solution method.It is found that the multiobjective ore blending model has become a trend,and the conventional indexes such as grade and ore quantity are mostly studied,while the unconventional indexes such as oxidation rate and rock type are not studied.To be consistent with the actual situation of the mine,based on the mine production scenario of multiple ore loading points and multiple ore unloading points,the polymetallic multi-objective ore blending model is built to improve the utilization rate of resources.The model minimizes the metal grade deviation,transportation work and ore-rock type ratio deviation,with oxidation rate,production capacity and total ore blending amount as constraints.(3)Aiming at the complex ore blending model,the improved NSGA-? algorithm is adopted to solve it.The motivation behind this is that the convergence of NSGA-? algorithm needs to be improved and matching pool mechanism is not conducive to the diversity of solving multi-constrained problems,a new environment selection mechanism and matching pool mechanism are designed.(4)The polymetallic multi-objective ore blending model and the improved NSGA-? algorithm were applied to an open pit mine in Henan.Combined with the actual production data,the reasonable ore blending scheme was obtained and a better ore blending plan was made according to various ore blending schemes in different cases.The multi-aspect analysis of ore blending results proves the effectiveness of multi-metal ore blending decision.