中文摘要：

研究MBR膜通量，进行膜污染预测，是当今污水处理研究领域的重要课题之一。为了有效，准确地预测MBR膜通量，提出一种改进的极限学习机( PSO-ELM)预测模型。极限学习机（ELM）能够有效地克服反向传播（BP）算法的缺陷，并能以极快的速度获得很好的泛化性能。由于随机给定输入权值和隐层阈值，ELM通常需要较多隐含层节点才能达到理想精度。利用粒子群算法（PSO）对极限学习机（ELM）的权值和阈值进行优化，建立PSO-ELM预测模型，将提取的主成分作为该模型的输入，膜通量作为模型输出。研究结果表明，该模型对MBR膜通量预测具有较好的泛化能力和更高的预测精度。

英文摘要：

MBR membrane flux research for membrane fouling prediction is one of the important topics on today's sewage treatment research field. In order to effectively and accurately predict the flux of MBR membrane, a prediction model was proposed based on improved extreme learning machine (PSO-ELM). Extreme Learning Machine could overcome the drawbacks of backpropagation algorithm with extreme learning speed and better generalization perform-ance. ELM usually requires more hidden layer nodes to achieve the desired accuracy because of the random input weights and hidden layer threshold. Using particle swarm optimization( PSO) to optimize the input weights and hidden layer threshold of extreme learning machine ( ELM), establish PSO-ELM prediction model, treat the extractive princi-pal components as the input of the prediction model, the membrane flux as the model output. The results show that the proposed model has better generalization ability and higher prediction precision for membrane flux.