Predictive Indicators Of Water In Wastewater Treatment Based On Improved BP Artificial Neural Network

Water is the source of all things, and people can’t live without water all the time. Our country is now lack of water resources relatively in the world. To protect water resources and prevent water pollution become a hot issue in today’s research. Sewage treatment plant plays an essential role in protecting the water environment. However, many sewage treatment plants through artificial laboratory detection of the daily sewage indicators, but it has certain hysteresis, like BOD5need to be known after five days, but the bad water has been discharged. In addition, in order to improve the water quality, requires a lot of manpower, time and money invested. In recent years, people will be soft measurement techniques used in the sewage treatment process, and control process in time, and to achieve the purpose of efficient sewage treatment.This thesis mainly establish forecasting model of sewage treatment plant of water index, and adjusted by changing the parameters, and analyze trend of water index change, and guide the practice of wastewater treatment plant.The main idea include:First, analyze and learn wastewater treatment process in Inner Mongolia region, understand the main process of wastewater treatment process and mechanism, study A2/O process of activated sludge. Second, analyze the situation of water quality and determine the input index of model, with Zhanggaiying sewage treatment plant as an example in huhhot. Third, the BP neural network model which has10input variables,13hidden neurons,1output variable. And training function is traingdx, learning function is learngdm, hidden and output layer is tansig and purelin. And adopted is0.01, momentum coefficient is0.04, the maximum training number20000, the goal is0.0005(TP and SS) and0.0003(BOD5). And model prediction results meet accuracy requriement.Fourth, it change the control parameters (DO, pH, water temperature), and the model carry the forecast of BOD5and TP, and analyzes the change trend which the predicted results of three indexes, to guide the sewage treatment process.