The Research On Soft Measurement And Prediction Modeling Of Effluent COD In Sewage Treatment

In recent years,with the national economic development,the problem of water pollution has become increasingly serious.Sewage management is already an urgent problem to be solved.Therefore,the research and application of sewage treatment system is of great significance.This paper takes the sewage treatment of an environmental technology company in Anhui Province as the research object,and studies the prediction of effluent COD parameters.The process of wastewater treatment is complex and the water changes are non-linear and unknown.The ASM-related model used in the traditional sewage treatment has increased the interference parameters of the measurement parameters because of its consideration of the parameters and processes,and finally brings the undesirable effect to the forecast model.Artificial neural network has better fault tolerance,adaptive ability and can carry out large-scale data parallel computing.It is suitable for multi-factor,multi-condition of the occasion,and dealing with data based on a fuzzy method.Therefore,it has some advantages to apply the neural network to the prediction of wastewater qualities.The main contents of this paper includes:Firstly,aiming at the advantages of soft sensor application,a soft sensor model of neural network should be established and simulated.The improved methods based on BP algorithm and Elman algorithm is added to the lack of model prediction effect,and the accuracy of the model is improved.Secondly,the basic principles of support vector machines,the selection methods of kernel functions and parameters are studied,and the support vector machine is applied to the prediction modeling of sewage.Two methods,i.e.,such as grid search optimization and particle swarm optimization,are used to optimize the parameters of the prediction model,and achieved good predictive effect.Finally,a prediction model based on the support vector machine and the improved neural network is established to deal with the contradiction data caused by the local minimum and the auxiliary variables in the modeling of the neural network.At the same time,in order to solve the problem of correlation between single neural networks in the combined model,the main component regression(PCR)is used to perform the data weighted fusion of the submodule output.In the end,the constructed soft-sensing model has a great improvement in prediction accuracy and has good robustness and generalization ability.The contribution of this paper is to establish a soft measurement effluent COD modelbased on the combination of support vector machine and improved neural network,and verify and compare the single neural network model and the support vector machine model to verify the excellent prediction effect of this method.More accurate prediction of effluent COD parameters.