生物炭吸附新兴污染物的机器学习模型预测与机制研究
屈云易 , 刘玲 , 冉文悦 , 宋潘宇 , 武峥 , 洪思奇
现代化工 ›› 2026, Vol. 46 ›› Issue (11) : 202 -207.
生物炭吸附新兴污染物的机器学习模型预测与机制研究
Machine learning model prediction and mechanism study of biochar adsorption for emerging pollutants
为提升新兴污染物治理效率,研究系统整合549组生物炭吸附实验数据,通过非参数插补方法处理缺失值,采用3种典型算法建立多维度特征生物炭吸附新兴污染物预测模型。其中,XGBoost模型表现出最优拟合性能(测试集R2=0.95,RMSE=8.32 mg/g)。特征重要性分析结果表明,表面氮官能团与污染物摩尔折射率共同构成吸附协同机制的核心要素,证实表面化学作用在吸附过程中的贡献超过物理作用;平均孔径的重要性显著高于比表面积,且仅当孔径范围与污染物分子尺寸匹配时,微孔填充效应才显著增强。明确表面化学性质在提高新兴污染物吸附量的作用更为关键,为高性能生物炭的定向制备及新兴污染物的高效去除提供了理论支撑。
To enhance the removal efficiency of emerging pollutants (EPs),this study systematically integrated 549 sets of biochar adsorption experimental data.Missing values were addressed using non-parametric imputation methods,and predictive models for biochar-based adsorption of EPs were developed by employing three typical algorithms with multi-dimensional features.Among them,the XGBoost model demonstrated superior fitting performance,achieving a coefficient of determination (R2) of 0.95 and a root mean square error (RMSE) of 8.32 mg/g on the test set.Feature importance analysis revealed that:surface nitrogen-containing functional groups and the molar refractivity of pollutants constitute the core elements of the synergistic adsorption mechanism,confirming that surface chemical interactions contribute more significantly to the adsorption process than physical interactions;the importance of mean pore size is notably higher than that of the specific surface area,with the micropore-filling effect being markedly enhanced only when the pore size range matches the molecular dimensions of the pollutants.This study clarifies that surface chemical properties play a more critical role in increasing the adsorption capacity of EPs,providing theoretical support for the tailored preparation of high-performance biochar and the efficient removal of emerging pollutants.
machine learning / biochar / adsorption mechanism / emerging pollutants
| [1] |
|
| [2] |
|
| [3] |
|
| [4] |
|
| [5] |
|
| [6] |
|
| [7] |
|
| [8] |
|
| [9] |
|
| [10] |
|
| [11] |
|
| [12] |
|
| [13] |
|
| [14] |
|
| [15] |
|
| [16] |
|
| [17] |
|
| [18] |
|
| [19] |
|
| [20] |
|
| [21] |
|
| [22] |
|
| [23] |
陈一飞, 张晓晴, 谭康豪, |
| [24] |
花文静, 张彦平, 王庆元, |
陕西省自然科学基础研究计划(2025JC-YBMS-158)
/
| 〈 |
|
〉 |