Modeling Of A~2O Process Based Urban Sewage Treatment Plant

With the increasing pollution of Water environment and shortage of water resource, sewage treatment needs to develope energetically.The establishment of a reliable system for a sewage treatment to morritor the important process parameters is necessary.First quality of the effluent water can be controled with specified standard.Second the energy consumption of the entire process can be optimized with energy saving and emission reduction.These two aspects are stand for two important parameters of sewage treatment process:the effluent COD and DO in biochemical pool.The main target of this paper is to build an operating guidance model both meet the effluent water quality under the premise of optimal DO value, achieve energy saving purpose.Way to establish the model for water treatment can be divided into the mechanism model and the intelligent control based method, the former needs to a deep understanding in water treatment process, the latter just needs the right sample input and output data. Because of the complexity on the sewage treatment process, this paper chooses the latter one.The research object of this paper is city sewage treatment system based on the A2O. The paper firstly analyze the A2O wastewater treatment process, the characteristics of study, and then use BP neural network and least squares support vector machine modeling and optimization on screening of processed data, and makes a comparison between the two methods.Finally, based on the understanding of above as well as the feasibility analysis of sewage treatment system modeling, put forward the multiple-model soft sensor modeling based on fuzzy C mean clustering. This method can use a variety of models to their respective advantages, accelerate data prediction speed, and improve the prediction accuracy and generalization ability.