基于PSO-Kriging模型的尾矿库三维稳定性分析  

Three-dimensional Stability Analysis of Tailings Pond Based on PSO-Kriging Model

作　　者：黄德镛[1] 黄日胜 史凯东 吕世玮 陈治宇 HUANG Deyong;HUANG Risheng;SHI Kaidong;LYU Shiwei;CHEN Zhiyu(Faculty of Land Resources Engineering,Kunming University of Science and Technology,Kunming 650093,China;Guizhou Energy Group Co.,Ltd.,Guiyang 550081,China)

机构地区：[1]昆明理工大学国土资源工程学院,昆明650093 [2]贵州能源集团有限公司,贵阳550081

出　　处：《有色金属(中英文)》2025年第3期474-483,共10页Nonferrous Metals

基　　金：国家自然科学基金资助项目(52174114)。

摘　　要：尾矿库是矿业活动中不可或缺的组成部分,同时也带来了显著的环境和安全风险。在尾矿累积过程中,物理沉淀、水动力作用和化学反应等多重因素导致尾矿的物理特性呈现出明显的空间差异性和分布不均匀性。现有的研究方法大多忽略了尾矿材料的不均一性和复杂性。而研究这些特性需进行大量的物理试验,虽然这些试验可以重复,但存在着系统误差且成本高昂,因此可以构建一个近似模型进行机械学习预测。以Kriging理论为基础,通过对多种寻优算法进行适应度对比以选出最佳的寻优算法,构建出高效的改进Kriging模型,为了验证PSO-Kriging模型的性能,采用估值分析与误差分析的方式对本模型插值效果进行综合评价,结果显示新模型提高了预测精度和变化趋势。在此基础上得到一组符合尾矿库实际情况的插值点特征力学参数。从空间变异性出发对尾矿库稳定性进行分析。基于一般沉积性数值模型,将插值点坐标与数值模型中网格模型中坐标相对应,通过Fish函数,将插值点力学参数即天然重度、黏聚力与内摩擦角的数据导入网格模型中。替换原网格点的力学参数,构建出考虑空间变异性的尾矿库三维数值模型,并对该模型进行分析,结果表明由于空间差异信息增加,计算结果更能反映实际状况。Tailings ponds are an indispensable part of mining activities,but they also bring significant environmental and safety risks.During the accumulation of tailings,multiple factors such as physical sedimentation,hydrodynamic action,and chemical reactions cause the physical properties of tailings to show obvious spatial differences and uneven distribution.Most existing research methods ignore the heterogeneity and complexity of tailings materials.To study these characteristics,a large number of physical experiments are required.Although these experiments can be repeated,they have systematic errors and are costly.Therefore,an approximate model can be constructed for mechanical learning prediction.Based on the Kriging theory,by comparing the fitness of multiple optimization algorithms to select the best one,an efficient improved Kriging model is constructed.To verify the performance of the PSO-Kriging model,a comprehensive evaluation of the interpolation effect of this model is conducted through value analysis and error analysis.The results show that the new model improves the prediction accuracy and trend.On this basis,a set of interpolation point characteristic mechanical parameters that conform to the actual situation of the tailings pond is obtained.The stability of the tailings pond is analyzed from the perspective of spatial variability.Based on the general sedimentary numerical model,the coordinates of the interpolation points are corresponded to the coordinates of the grid model in the numerical model.Through the Fish function,the mechanical parameters of the interpolation points,namely the natural unit weight,cohesion and internal friction angle,are imported into the grid model.The mechanical parameters of the original grid points are replaced to construct a three-dimensional numerical model of the tailings pond considering spatial variability,and the model is analyzed.The results show that due to the increase of spatial difference information,the calculation results can better reflect the actual situation.

关 键 词：尾矿库 稳定性分析 KRIGING插值 PSO优化算法 三维数值模拟 

分 类 号：TD928[矿业工程—选矿]