The Application Of Artificial Neural Network On Water Quality Evaluation And Prediction In Chao Lake

In this paper, The author studies on the application of artificial neural network on water quality evaluation and prediction in Chao Lake.On ware quality evaluation ,according to the need of evaluating the nutritious levels of Chao Lake , TP, TN, COD_Mn and water temperature are used as the evaluation indexes, the standards of the nutritious levels evaluation are confirmed, Back-Propagation(BP) network model and Back-Propagation network in series with degree of membership(BPDM) model are developed, using the models, a evaluation was implemented for 6 points in Chao Lake from Spring, year 2000 to Winter, year 2002. By the evaluation, the nutritious states in every season during the three years of the 6 points are getted, and we can conclude that as a whole, the lake belongs to medium to eutrophication level, with the western part of eutrophication level and the eastern part of medium level.On water quality prediction, the relation between the concentration of chl-a and the eutrophication states is established,as the reflection indexes,the water temperature and the concentration of chl-a are identified. Using neural network model, the trend of eutrophication of the next month is predicted. Gaining sufficient samples with neural network inserted method, the trend of eutrophication of the next week is predicted. The results shows that water temperature and chl-a effect the concentration of chl-a of next week very well.The network using them as input variables is simple and prompt. Using neural network, the short term trend of eutrophication effectively can be predicted.Finally,a GM(1,1) model in series with neural network is developed, by calculates with the monitoring data of Chao Lake, we can prove that the model will make the grey model prediction accurately.