| Water is the source of life, but the water resources in the world are suffering from different degrees of pollution. Wastewater treatment is the main way to solve the water pollution.The pre-denitrification biofilm sewage treatment system consists of the BAF and pre-denitrification biofilm technology. It can not only effectively degrade organic matter, but also to remove the amount of nitrogen in wastewater.However,the aeration system energy consumption is big and high operation cost.Predictive modeling of effluent COD density is useful to adaptive adjustment of the control system, for one thing, timely adjustment of fans’ rotational speed, so as to decrease the energy consumption and save money.Relies on the major scientific and technological projects in Anhui Province which was jointly sponsored by the paper and some environmental technology company, the essay establishs the predictive model of effluent COD density which based on the ABAF / OBAF Pre biofilm denitrification wastewater treatment process and application project.On one hand, the organic degradation model was built according to Monod model and the principle of conservation of materials. Parameters in model were set by the field data collected and empirical value. It consider the influence of uncertainties,the random process theory was applied to the model. Then the corresponding differential equation is built and solved.On the other hand, according to the data fusion theory about Bayesian theory and Maximum Likelihood Estimation the other prediction model of effluent COD density based on the Gaussian Process was built, and the initial value of input parameters were set by PCA. After that, it verify the prediction model results with Matlab software and programming examp. Finally, the essay compare the above two models of forecasting results, it concluded that the gaussian process prediction model based on data fusion of prediction effect is more close to the actual measured values. |
PDF Full Text Request