论文部分内容阅读
在污水处理过程中,出水总磷(Total Phosphorus,TP)是衡量污水处理效果的关键参数之一。本文针对目前出水TP难以实时测量的问题,提出了一种基于模糊神经网络(FNN)的出水TP软测量方法。该软测量方法通过实际运行数据,利用偏最小二乘法(Partial Least Squares,PLS)筛选出与出水TP相关性强的过程变量;同时,利用FNN建立了出水TP与相关性变量之间的软测量模型,并将该方法嵌入到污水处理运行系统。实验结果显示该软测量方法能够实现出水TP的实时预测,并且具有较好的预测精度。
During the wastewater treatment, total phosphorus (TP) is one of the key parameters to measure the effectiveness of wastewater treatment. In this paper, aiming at the problem that TP is difficult to be measured in real time, a soft TP measurement method based on FNN is proposed. The soft-sensing method is based on PLS to select process variables that have strong relativity with effluent TP through actual operating data. At the same time, soft-sensing between TP and correlation variables is established by FNN Model, and embed the method into the sewage treatment operation system. The experimental results show that the soft-sensing method can realize real-time prediction of TP and has good prediction accuracy.