地下金属矿山采矿成本预测模型  

Mining Cost Prediction Model for Underground Metal Mine

作　　者：李国清[1] 吴炳书 侯杰[1] 王浩 王进强[1] 陈连韫 范纯超 LI Guoqing;WU Bingshu;HOU Jie;WANG Hao;WANG Jinqiang;CHEN Lianyun;FAN Chunchao(School of Civil and Resource Engineering,University of Science and Technology Beijing,Beijing 100083,China;Shandong Gold Group Co.,Ltd.,Jinan 250102,China)

机构地区：[1]北京科技大学土木与资源工程学院,北京100083 [2]山东黄金集团有限公司,山东济南250102

出　　处：《金属矿山》2022年第5期62-69,共8页Metal Mine

基　　金：国家自然科学基金项目(编号:52074022)。

摘　　要：为了解决现代矿山采矿成本的核心构成发生变化、成本管控难度增加等问题,深入剖析了采矿成本的关键影响因素,基于PCA-BP神经网络构建了成本预测模型,形成了适用于现代矿山的成本核算方法。针对机械化开采、开采深度增加、工况条件差异所带来的采矿成本变化,分析了现代矿山企业采矿作业成本的构成特点,以此为基础运用作业成本法精细化核算各个采矿作业单元的成本费用。总结成本变化规律发现,采矿成本主要受到采场作业空间大小、温度、深度、运输距离以及工人工作效率、工人工作经验、设备铲运效率、设备服务年限、燃油消耗率、油料消耗率10个关键因素影响,在采用主成分分析法对影响因素进行降维后,提取主成分作为成本预测变量,运用PCA-BP神经网络构建了成本预测模型。采用山东某地下金属矿山的成本数据对该模型进行了训练和验证,预测值与实际值的平均相对误差为3.80%。研究表明:所建模型的预测结果可靠、精度满足要求,可以为现代矿山企业作业成本控制、成本计划制定提供依据。In order to solve the problems of the core components of modern mine mining costs have changed and difficulty of cost control have increased,analyzed the key influencing factors of mining costs,constructed a cost prediction model based on PCA-BP neural network,formed a cost accounting method that is suited to modern mines.In response to the changes in mining costs caused by mechanized mining,increased mining depths,and differences in working conditions,obtained the characteristics of mining cost components of modern mining enterprises,based on this,applied the job costing method to account for the costs of each mining unit in detail.By summarizing the change pattern of mining cost,it was concluded that the mining cost is mainly influenced by 10 key index factors,such as the size of operating space,temperature,depth and transportation distance of the quarry,as well as workers′efficiency,working experience,equipment shoveling efficiency,service life of equipment,fuel consumption rate and oil consumption rate,using the principal component analysis method to reduce the dimensionality of the influencing factors,extracted the principal components as cost prediction variables,and established the cost prediction model based on PCA-BP neural network.Taking the cost data of an underground metal mine in Shandong Province as an example,the model is trained and validated,and the prediction results were also analyzed and evaluated.The results show that the average relative error between the predicted and actual values of PCA-BP neural network is 3.80%,which indicates that the model has good prediction effect and high accuracy,which provides a basis for cost control and cost planning of modern mining enterprises.

关 键 词：地下金属矿山 采矿成本 成本预测 预测模型 成本影响因素 

分 类 号：TD853[矿业工程—金属矿开采]