中文摘要：

曝气是MBR膜污染的操作条件影响因子中的一个重要参数,曝气强度过大易造成膜丝断裂,过小又不能减缓膜污染。针对该问题,本研究首先运用灰色模型对中空纤维膜不同使用阶段中的最佳曝气强度值进行粗略预测。再将影响膜过滤性能的三个因素作为BP神经网络的输入,不同膜清洗次数后的最佳曝气强度作为输出,进行曝气的BP网络模型预测。最后将灰色模型的预测值及影响膜过滤性能的三个因素作为灰色神经网络的输入,最佳曝气强度作为输出,进行曝气的灰色神经网络预测。通过对两个神经网络模型的预测结果对比分析,得出结论灰色神经网络模型优于BP神经网络模型。

英文摘要：

Aeration is an important parameter in the operating conditions of MBR membrane pollution,and the aeration intensity is too large to cause the rupture of membrane,which is too small to slow down the membrane fouling.To solve this problem,this study first uses the gray model to predict the best aeration intensity value of hollow fiber membrane in different stages.Three factors which affect the membrane filtration performance as the input of the BP neural network,the best aeration intensity of different membrane cleaning times as the output,the BP network model prediction.Finally,the grey model＇s predictive value and the three factors which affect the filtration performance of the membrane are used as the input of the grey neural network,the best aeration intensity is the output,and the grey neural network prediction.After comparing the prediction results of two neural network models,it is concluded that the grey neural network model is better than the BP neural network model.