【摘要】 快速准确判别矿井涌(突)水水源对保障矿井安全生产有重要意义。近年来人类的活动对不同含水层的影响与日俱增,为提高矿井涌(突)水水源判别的准确性,提出选取地下水中7种常见离子浓度,和能够充分反映人类活动痕迹的硝酸根离子浓度及化学需氧量作为水化学判别指标,采用随机森林模型进行矿井涌(突)水水源判别。为验证选取指标和判别方法的有效性,以大孤山铁矿为例,将数据输入随机森林模型进行100次交叉验证,并将验证结果与支持向量机模型和极限学习机模型进行比较。结果表明,随机森林模型预测结果稳定性较强,预测正确率不容易波动;随机森林在建模过程中参数拥有宽广的适应范围。树的棵数为50时,训练误差趋于稳定,改变树的棵数对预测结果没有实际影响,而其余二者对参数选取较为敏感;随机森林的参数可以通过袋外数据(OOB)错误率简单地得到,而其余二者参数调整时需要通过交叉验证的方式才可以取得;随机森林对训练样本进行验证,正确率可达100%,对测试样本进行验证,正确率可达97.38%,两项精度均优于支持向量机与极限学习机;随机森林模型拥有更高的预测精度和鲁棒性,在矿井涌(突)水水源判定方面有较好的应用前景。更多还原

【Abstract】 Quick and accurate detection of mine water source is of great importance for the safety of mine production. In recent years, the impact of human activities on different aquifers has increased. In order to improve the accuracy of water source in mine water, seven common ion concentrations in groundwater were selected, as well as the nitrate ion concentration and the chemical oxygen demand. The water chemical discriminant index could fully reflect the traces of human activity. And then the random forest model was used to identify the mine water source. The Dagushan iron ore was taken as an example to verify the validity of the selected indicators and methods. The data was input into the random forest model for 100 cross-validation, and the verification results were compared with the support vector machine model and the extreme learning machine model. The results of the random forest model are stable and the prediction accuracy is not susceptible to fluctuation. The random forest model has a wide range of adaptation parameters in the modeling process. When the number of trees is 50, the training error tends to be stable, and changing the number of trees has no practical impact on the prediction results, while the other two models are more sensitive to parameter selection. The parameters of the random forest model can be obtained simply by the out-of-bag error rate, but the other two model parameters need to be obtained through cross-validation. The random forest model verifies the training samples, and the correct rate can reach 100%. The random forest model also verifies the test samples, and the correct rate can reach 97.38%. The accuracy of the two correct rates is better than the support vector machine and the extreme learning machine. The random forest model has higher prediction accuracy and robustness, and has a better application prospect in the determination of mine water source.更多还原