使用聚胺薄膜膜进行增强的海水淡化,使用组合ML化学测量方法和SHAP分析
Jamilu Usman1, Sani I Abba2,3, Fahad Jibrin Abdu4
1Interdisciplinary Research Centre for Membranes and Water Security (IRC-MWS), King Fahd University of Petroleum and Minerals Dhahran 31261 Saudi Arabia saniisaabba86@gmail.com.
RSC advances
|October 3, 2024
概括
机器学习模型,特别是合集,准确地预测了海水淡化膜的性能 (平均水流量和盐排放). SHAP分析揭示了MDP和TMC等影响效率的关键特征,为可持续海水淡化铺平了道路.
科学领域:
- 材料科学与工程 材料科学与工程
- 化学工程是化学工程的重要组成部分.
- 环境科学 环境科学
背景情况:
- 全球淡水短缺需要先进的海水淡化技术.
- 薄膜复合聚胺膜对于高效的海水淡化至关重要.
- 预测建模对于优化膜性能指标至关重要,例如平均水流量 (AWF) 和平均盐排斥率 (ASR).
研究的目的:
- 应用机器学习 (ML) 算法和组合技术来预测和增强淡化膜中的AWF和ASR.
- 使用夏普利添加式解释 (SHAP) 进行模型解释性和特征重要性分析.
- 为了证明整体ML模型的优越性,而不是单个算法,以优化海水淡化过程.
主要方法:
- 使用的机器学习算法:支持矢量机 (SVM),神经网络 (NN),线性回归 (LR) 和多变量线性回归 (MLR).
- 利用组合技术 (例如,NN-E,LR-E) 来提高AWF和ASR的预测准确性.
- 应用SHAP分析以确定影响膜性能的关键特征.
主要成果:
- 神经网络 (NN) 显示出强大的个体性能,但组合模型 (NN-E,LR-E) 实现了高超的预测准确性,近乎完美的纳什-萨特克利夫效率 (NSE).
- 与单个模型相比,集成模型在预测AWF和ASR方面显著减少了错误 (MAE,RMSE).
- SHAP分析确定了膜冷却 (MDP) 和1,3,5-三二 (TMC) 作为关键性能驱动因素,MDP显著影响ASR.
结论:
- 集成ML方法,特别是NN-E和LR-E,为预测海水淡化膜效率 (AWF,ASR) 提供了更高的精度.
- SHAP分析为特征的重要性提供了关键的见解,使膜设计和操作能够有针对性的改进.
- 整合智能ML组合和SHAP可解释性可以推进可持续的海水淡化,优化资源使用并最大限度地减少对环境的影响.
相关概念视频
Dialysis
610
Dialysis is a diffusion-based purification process that separates analyte molecules from a complex matrix. This is accomplished by allowing molecules in the solution to pass through a semipermeable membrane into a liquid on the other side. The membrane is usually made of cellulose acetate or cellulose nitrate, and the second liquid must be miscible with the solution. Ions (e.g., chloride or sodium) or organic molecules (e.g., glucose) can pass through the membrane pores, which generally have...
610
Potentiometry: Membrane Electrodes
484
Membrane electrodes, also known as p-ion electrodes, use membranes that selectively interact with free analyte ions, generating a potential difference across the membrane. The resulting membrane potential, known as the asymmetry potential, is not zero even when analyte concentrations on both sides of the membrane are equal. The membrane's response is typically not selective to a single analyte but proportional to the concentration of all ions in the sample solution capable of interacting at...
484


