| 随机森林模型预测岩溶区酸性煤矿井水锰污染 |
| |
| 引用本文： | 李冲.随机森林模型预测岩溶区酸性煤矿井水锰污染[J].中国煤炭地质,2021(3):43-47,59. |
| |
| 作者姓名： | 李冲 |
| |
| 作者单位： | 中国煤炭地质总局水文地质局 |
| |
| 基金项目： | 凯里市鱼洞河流域环境综合治理工程可行性研究(凯发2018(0366号))。 |
| |
| 摘 要： | 酸性煤矿井水严重威胁地下水的水质。如何更有效对受影响区域的地下水源进行动态监测是当前的一个关键问题。采用随机森林中的回归模型,利用自变量(采空区水位、岩溶水位、pH值、泉水流量、电导率)和因变量(污染离子浓度)的相关性,建立回归模型;使用测试数据进行误差分析,结果证明模型准度较高,所得预测值具有参考价值;得出各自变量对因变量影响的重要程度,分析结果与实际情况相符合。试验表明,随机森林回归模型在酸性煤矿井水污染预测方面具有适用性,可作为辅助手段监测水质污染情况,对今后工作有一定的指导意义和经济价值。
|
| 关 键 词： | 酸性煤矿井水 地下水污染 水质监测 随机森林 |
Prediction of Karst Region Acidic Coalmine Water Manganese Pollution Based on Random Forest |
| |
| Li Chong.Prediction of Karst Region Acidic Coalmine Water Manganese Pollution Based on Random Forest[J].Coal Geology of China,2021(3):43-47,59. |
| |
| Authors: | Li Chong |
| |
| Institution: | (Hydrogeological Exploration Bureau,CNACG,Handan,Hebei 056004) |
| |
| Abstract: | Acidic coalmine water has seriously threatened groundwater quality.How to carry out groundwater dynamic monitoring in the impacted areas is the key issue at present.Based on the regression model in Random Forest,through interdependency between arguments(gob area water-level,karstic water-level,pH,spring water flow,electric conductivity)and dependent variable(contaminating ion concentration)has established regression model.Using tested data have carried out error analysis;the results proved the model has higher accuracy,predicted data have reference value;level of importance from arguments to dependent variable worked out,the analyzed results in accordance with the realities.The test has shown that the Random Forest regression model is adaptable to acidic coalmine water pollution prediction;it can be an auxiliary means in water quality pollution monitoring.The study has certain guidance and economic values for works henceforth. |
| |
| Keywords: | acidic coalmine water groundwater pollution water quality monitoring Random Forest |
| 本文献已被 CNKI 维普 等数据库收录！ |
|
