| 引用本文: | 陈君岩,周志伟,李星,赵元添,李晓滢,王天阳.可解释深度学习识别膜污染影响因子[J].哈尔滨工业大学学报,2026,58(6):101.DOI:10.11918/202507019 |
| CHEN Junyan,ZHOU Zhiwei,LI Xing,ZHAO Yuantian,LI Xiaoying,WANG Tianyang.Interpretable deep learning for membrane fouling factor identification[J].Journal of Harbin Institute of Technology,2026,58(6):101.DOI:10.11918/202507019 |
|
| |
|
|
| 本文已被：浏览 891次 下载 78次 |
码上扫一扫！ |
|
|
| 可解释深度学习识别膜污染影响因子 |
|
陈君岩,周志伟,李星,赵元添,李晓滢,王天阳
|
|
(北京工业大学 建筑工程学院,北京 100124)
|
|
| 摘要: |
| 膜污染严重制约超滤（UF）技术在城市污水处理中的应用,成分复杂的二级出水有机物（EfOM）是其主要诱因,但关键膜污染组分尚缺乏有效的识别方法。为此,基于TabPFN（Tabular Prior-Data Fitted Network）深度学习模型,结合SHAP（SHapley Additive exPlanations）方法,构建关键膜污染因子识别框架。以改性陶瓷超滤膜（MCUM）2 h运行周期内的水质与膜污染数据为基础,选取UV254、溶解性有机碳（DOC）、荧光区域积分值及荧光组分强度（Fmax）为输入,预测末端比通量、可逆阻力与不可逆阻力。结果显示,TabPFN模型无需超参数调整即可实现高精度预测（末端比通量R2=0.97）,整体性能优于传统机器学习模型。进一步引入后验集成（PHE）策略,预测性能提升至R2=0.98。SHAP分析表明,色氨酸类、富里酸类和腐殖酸类物质分别在膜通量下降、不可逆阻力和可逆阻力的形成中具有主要影响。二维偏依赖图（PDPs）结果进一步揭示了这些组分之间显著的非线性协同作用,特别是富里酸类与芳香性物质在不可逆污染过程中的协同增强效应。本研究揭示了EfOM中主导膜污染的关键组分,为膜污染控制策略优化与污染机制理解提供了理论支撑。 |
| 关键词: 膜污染 二级出水有机物 深度学习 可解释性分析 TabPFN模型 |
| DOI：10.11918/202507019 |
| 分类号:X703.1 |
| 文献标识码:A |
| 基金项目:国家自然科学基金(52370022) |
|
| Interpretable deep learning for membrane fouling factor identification |
|
CHEN Junyan,ZHOU Zhiwei,LI Xing,ZHAO Yuantian,LI Xiaoying,WANG Tianyang
|
|
(The College of Architecture and Civil Engineering, Beijing University of Technology, Beijing 100124, China)
|
| Abstract: |
| Membrane fouling seriously restricts the application of ultrafiltration (UF) technology in municipal wastewater treatment. Secondary effluent organic matter (EfOM), characterized by its complex composition, is considered the primary cause of membrane fouling. However, there is no method to effectively identify the key components of membrane fouling. To address this, this paper proposed a key membrane fouling factor identification framework based on the tabular prior-data fitted network (TabPFN) deep learning model by using SHapley Additive exPlanations (SHAP) method. Water quality and membrane fouling data were collected from a modified ceramic ultrafiltration membrane (MCUM) during a two-hour operation cycle. UV254, dissolved organic carbon (DOC), fluorescence regional integration values, and fluorescence component intensities (Fmax) were selected as input variables to predict normalized specific flux, reversible resistance, and irreversible resistance. The results show that the TabPFN model achieves high-precision prediction without hyperparameter tuning (R2=0.97 for normalized specific flux), outperforming conventional machine learning models in terms of overall performance. Further improvement is obtained by introducing a post-hoc ensembling (PHE) strategy, increasing the prediction performance of R2 to 0.98. SHAP analysis indicates that tryptophan-like, fulvic acid-like, and humic acid-like substances have predominant influences on the reduction of membrane flux and the formation of irreversible resistance and reversible resistance, respectively. PDPs results further reveal significant nonlinear synergistic effects among these components, especially the synergistic enhancement effect of the fulvic acid-like and aromatic substances in the irreversible fouling process. This study reveals the key components of secondary EfOM responsible for membrane fouling, providing theoretical support for membrane fouling control strategy optimization and the understanding of fouling mechanisms. |
| Key words: membrane fouling secondary effluent organic matter deep learning interpretable analysis TabPFN model |
|
|
|
|
