Research On The Application Of Intelligent Information Processing In The Oxidation Ditch System

The oxidation ditch process is a realization of sewage purification technique. Simulating the correlation of oxidation ditch water quality parameters correctly is an important topic in the real-time control of on-line water treatment system.Artificial neural networks have been widely used in sewage purification system simulation and prediction as a promising approach because of their distinguishing features:self-organized,self-adaptive, error-contained and parallel,however the application is still in primary state,and the optimization of neural network structure and its performance is not considered too much, which have a great effect on the precision of simulating results.In this thesis,oxidation ditch simulation is investigated,and the solutions to the optimization improvement of oxidation ditch system are given.The results are as follows:(1) The methods of oxidation ditch modeling are reviewed;the advantages and disadvantages of existing methods are summarized.(2) The BP neural network model for effluent SS,COD simulation is built,the precision of effluent TN,TP simulation using BP neural network is tested,and the result is not agreeable,so a combination model using PCA,k-means algorithm and RBF neural network is built for TN,TP modeling.(3) The improved immune algorithm is combined with PCA and RBF neural network for effluent TN,TP modeling.(4) Based on models having been built,the correlation of oxidation ditch water quality parameters is analyzed.