基于降雨量数据的程潮铁矿涌水量时序性预测模型  

Temporal Prediction Model of Water Inflow in Chengchao Iron Mine Based on Rainfall Data

作　　者：范明星 任高峰[1,2,3] 吴文博 鲁习奎 李吉民 张聪瑞[1,2,3] FAN Mingxing;REN Gaofeng;WU Wenbo;LU Xikui;LI Jimin;ZHANG Congrui(School of Resources and Environmental Engineering,Wuhan University of Technology,Wuhan 430070,China;Key Laboratory of Green Utilization of Key Non-metallic Mineral Resources,Ministry of Education,Wuhan 430070,China;Key Laboratory of Mineral Resources Processing and Environment of Hubei Province,Wuhan 430070,China;Jingmen Natural Resources and Planning Bureau,Jingmen 448000,China;Chengchao Mining Company of WISCO Resources Group,Ezhou 436051,China)

机构地区：[1]武汉理工大学资源与环境工程学院,湖北武汉430070 [2]关键非金属矿产资源绿色利用教育部重点实验室,湖北武汉430070 [3]矿物资源加工与环境湖北省重点实验室,湖北武汉430070 [4]荆门市自然资源和规划局,湖北荆门448000 [5]武钢资源集团程潮矿业有限公司,湖北鄂州436051

出　　处：《金属矿山》2024年第6期212-219,共8页Metal Mine

摘　　要：无底柱分段崩落法开采必然形成大范围的地表塌陷区,所分布的导水裂隙将导致大量的地表降水向下渗透,引起巷道涌水量骤增,程潮铁矿现已进入-500 m开采水平,大气降雨与崩落法开采所引发的裂隙持续动态发展之间的相互作用机制十分复杂。因此,为科学准确地预测井下涌水量,本研究提出了一种以降雨量为输入数据的涌水量灰色GM(1,2)时序性预测模型。以程潮铁矿2019—2021年的降雨量、涌水量实际数据为训练样本,充分考虑崩落法开采对上覆岩体的持续影响,引入时序性系数K对模型进行优化,最终建立的灰色GM(1,2)时序性预测模型与传统的GM(1,2)预测模型相比,预测精度平均提高了7.79%。运用该模型对2022年的涌水量进行预测检验结果表明,其旱、雨两季的预测精度分别为93.51%、93.58%,预测效果较好。研究成果是通过地表降雨量数据直接预测矿山井下涌水量数据的一种有效方法。The non-pillar sublevel caving mining will inevitably form a large-scale surface subsidence area.The distributed water-conducting fractures will lead to a large amount of surface precipitation infiltrating downward,causing a sudden increase in roadway water inflow.Chengchao Iron Mine has now entered the mining level of-500 m.The interaction mechanism between atmospheric rainfall and the continuous dynamic development of fractures caused by caving mining is very complicated.Therefore,in order to predict the underground water inflow scientifically and accurately,this paper proposes a grey GM(1,2)time series prediction model of water inflow with rainfall as input data.Taking the actual data of rainfall and water inflow in Chengchao Iron Mine from 2019 to 2021 as training samples,fully considering the continuous impact of caving mining on overlying rock mass,the time series coefficient K is introduced to optimize the model.Finally,the grey GM(1,2)time series prediction model is established.Compared with the traditional GM(1,2)prediction model,the prediction accuracy is improved by 7.79%on average.The model is used to predict the water inflow in 2022.The results show that the prediction accuracy of dry and rainy seasons is 93.51%and 93.58%respectively,and the prediction effect is good.The research results are an effective method to directly predict the mine water inflow data through the surface rainfall data.

关 键 词：无底柱分段崩落法 涌水量预测 灰色系统理论 GM(1 2)模型 时序性系数 

分 类 号：TD742[矿业工程—矿井通风与安全]