Membrane wetting is a bottleneck that limits the widespread application of membrane distillation (MD) technologies. However, the prediction of membrane wetting is difficult, due to its unpredictable behavior with the chemical species in feed waters. We used response surface methodology (RSM) and artificial neural networks (ANN) to predict the wetting phenomena in direct contact membrane distillation (DCMD) for the treatment of synthetic wastewater. Experiments were performed at various concentrations of NaCl, CaSO4, humic acid, alginate, and sodium dodecyl sulfate (SDS) to examine their effects on the wetting. The RSM and ANN models were established using the experimental data and statistically validated by the analysis of variance (ANOVA). The results showed that both RSM and ANN are able to predict the time of wetting and recovery for the range of input variables. The model predictions suggested that the concentration of NaCl and SDS has the greatest influence on the prediction parameters. When the concentration of SDS was less than 5 mg/L, the concentration of NaCl was the dominant role in the wetting. On the other hand, the concentration of SDS was the predominant factor when the concentration of SDS was higher than 5 mg/L.

中文翻译：

---

表面活性剂对膜蒸馏膜结垢润湿的影响：响应面法 (RSM) 和人工神经网络 (ANN) 的应用

膜润湿是限制膜蒸馏 (MD) 技术广泛应用的瓶颈。然而，膜润湿的预测是困难的，因为它与进水中的化学物质的行为不可预测。我们使用响应面方法 (RSM) 和人工神经网络 (ANN) 来预测直接接触膜蒸馏 (DCMD) 处理合成废水中的润湿现象。在不同浓度的 NaCl、CaSO4、腐殖酸、藻酸盐和十二烷基硫酸钠 (SDS) 下进行实验，以检查它们对润湿的影响。使用实验数据建立 RSM 和 ANN 模型，并通过方差分析 (ANOVA) 进行统计验证。结果表明，RSM 和 ANN 都能够预测输入变量范围内的润湿和恢复时间。模型预测表明，NaCl 和 SDS 的浓度对预测参数的影响最大。当 SDS 浓度小于 5 mg/L 时，NaCl 浓度在润湿中起主导作用。另一方面，当 SDS 浓度高于 5 mg/L 时，SDS 浓度是主要因素。