Research of COD Concentration Prediction Model for River Monitoring Section Based on Artificial Neural Network

According to the diversity and nonlinear characteristics of effect factors for COD concentration, a model of predicting COD concentration for surface water environment monitoring section was established based on the BP Neural Network technology. The model was applied to predict COD concentration for Xiaojiemiao monitoring section of Taizi river in Anshan area in this paper. The results showed that the model could successfully predict the COD concentration of this river section. Moreover, the prediction model was capable of generalization performance so as to provide the new method for COD concentration prediction.