The Study of Prediction of Agricultural Source Pollution Emissions of Ammonia Nitrogen (AN) and Chemical Oxygen Demand (COD) Based on Bionic Neural Network Algorithm

Abstract

Nearly 40% of Chinese water pollution comes from agricultural sources of pollution, and the annual emissions are difference. If we want to control pollution emissions effectively, we need to accurately predict the amount of agricultural emissions of Ammonia Nitrogen (AN) and Chemical Oxygen Demand (COD). Due to the complex mechanism of the agricultural non-point source pollution, its emissions are very difficult to measure. Currently, the Bionics Research is in a stage of rapid development, and it continues to expand into many new areas of research. So the comprehensive study of Bionics and pollutant control study will be a good choice. This research used bionic BP（Back Propagation） neural network algorithm, and used pollution census data from 2002 to 2007 and established neural network model with neural network algorithm. And we predicted the agricultural sources of emissions of AN and COD with the data from 2008 to 2010. Finally we compared the predicted value and the actual value. Research results showed that, with using the bionic BP neural network, agricultural sources emissions of AN and COD are evaluated actually and the results indicate that the average error is under 5.0%. Research results proved that the model is effective. The neural network is a scientific predict method for the agricultural sources emissions of AN and COD. It can be widely used in the prediction of agricultural sources emissions of AN and COD.