A Prediction Of Pollution Of Karast Leakage In Baiji Ardealite Site

With the space of development more and more quickly in west of China, also, Karst groundwater pollution problems in GuiZhou province are getting more and more prominent. Especially, aluminium industry and phosphorus chemical industry mushroom in GuiZhou province, which produce much waste residue. In residue site, the leaching of waster seep through the land. Groundwater pollution has become a threat to the groundwater resources of Karst area.The paper analyses the geology and geographic environment of the ardealite site. Based on evaluating and analyzing the karst groundwater leakage of the slag yard, establishing a model of predicting the Krast groundwater seepage with BP Neural Network System with NNTool of neural network tool box in Matlab.The model is estiblished based on the relationship between rainfall, quantity of seepage and pollutant concentration in the site of seepage, adopting two types of backpropagation algorithm,that are resilient back-PROPagation and variable learnling rate backpropagation, and two means of data preprocessing. Chosse the best modle by adjusting parameter of models continually. At last compared with 1-D mathematical model to relize the feasibility of the BP neural networks model in karst groundwater.The BP neural network model is only with an eye to the relationship of datas import and outport, but neglecting the inherent mechanism, therefore the model simplify the process of movment of groundwater in Karst area. It offers reasons for controling karst groundwater leakage of the slag yard .