The Application Of BP Neural Network In UASB On Controlling And Operation

High-organic wastewater is frequently discharged nowadays, which is the characterized of with high content of contamination, heavy harmful, hard to be disposed with high costs. A BP neural network model is to be established to predict the effluent CODcr which he tannery water was selected to be experiment object. The anaerobic biological processes is to be managed and controlled from the model,Anaerobic biological processes are becoming an attractive alternative for treatment of industrial wastewater because of their performance advantages.However, the start-up and control of these processes are relatively complex because of the low specific growth rate of their microorganisms and their high sensitivity to variations in environmental conditions.Tt is difficult to quantitatively describe the anaerobic processes for numerous comprehensive mathematical models. Moreover the influent quantity, pH,alkalinity and the effluent volatile fatty acid all will have influence on the effluent quantity,the realition between each of them and the COD_cr is not linear.So the trational ways can't analyze the non-linear complicated anaerobe biological process.The paper adopts artificial neural network as analyzing tools.It has self-adaptive,self-learning and faulty-tolerant capacity as well as large-scale concurrent,which is particularly applicable for inaccurate and fuzzy information processing with many factors and conditions involved.The paper adopts visual basic6.0 as the computational platform.Base on monitoring date of sewage disposal station, the predicting values of the BP model are well accord with the monitoring data.The relative errors are almost less than 5% and mean value of relative errors is 3.28%. The model can accurately predict the effluent COD_cr-From the model the anaerobic biological processes is to be analyzed, the strategy and scheme to the problem are to be introduced.