基于BP神经网络的SBR法污水处理DO预测模型研究

Study on the DO forecasting model during wastewater treatment based on BP neural network

通过对某污水处理厂的SBR法污水工艺进行分析,得知其过程具有多变量、非线性、时变性与随机性,能否及时检测出DO值至关重要.然而DO的在线检测仪一般有一定的滞后,为此本文采用BP神经网络来建立DO的预测模型并通过MATLAB仿真研究,其结果表明该模型具有较好的预测效果.

This paper analyzes wastewater treatment technique by using SBR in a sewage treatment plant,then the process is informed that it is multi-variable,nonlinear,time-varying and random,moreover,the detection of DO is crucial in time.However,the online DO detector in general has a certain lag,therefore this paper use the BP neural network to build the prediction model of DO.The result shows that the model has good prediction through MATLAB simulation study.