Research And Application Of Prediction Algorithm For Abnormal Conditions Of Sewage Implementation

Sewage treatment is a non-linear and strong interference industrial process,so the causes of abnormal working conditions are also very complex.If the abnormal working conditions cannot be handled in time,a lot of losses and serious problems will be caused.Predicting abnormal conditions by prediction algorithm,the cause of abnormal conditions can be found in advance,and the occurrence of abnormal conditions can be prevented at a low cost.However,because of the complexity of sewage treatment,the traditional prediction algorithm can not predict the wastewater treatment process effectively,so the characteristics of the sewage data set should be considered in the study of the prediction algorithm.This paper takes sewage treatment as the application background,and the main research contents are as follows:(1)In view of the complex characteristics of high dimensional imbalance in sewage data set,an improved density peak algorithm(DPC-G)is proposed,and the simulation comparison experiment is carried out with four traditional algorithms.Through the analysis of the experimental results,it is found that the quality evaluation indexes of the cluster quality of DPC-G clustering such as ARI,FMI,AMI,CH,DB and Si are better than the control algorithm.(2)Based on the theory of single hidden layer feedforward neural network,the classification cost regularized learning machine is selected as the weak classifier.Combined with smote oversampling technology and bagging ensemble algorithm,a smote oversampling bagging ensemble algorithm based on classification cost regularized extreme learning machine is proposed＿ELM＿SMOTE＿Bagging)。 The experimental results show that the algorithm has higher performance in the classification of sewage data sets.(3)A warning software for abnormal working conditions of sewage treatment is designed by C # which includes user management,abnormal prediction and data management.After testing,the software is simple to use,stable and reliable,greatly improving the efficiency of staff.