Removal Of Algae By Pre-oxidation Enhanced Coagulation And Prediction Of Algae Removal Rate Based On Deep Learning

With the rapid development of China’s economy in recent years,a large number of industrial and agricultural wastewater is discharged into the water body,resulting in the eutrophication of some rivers,and then causing a series of algae outbreak problems.Algae outbreaks not only affect the ecological balance,but can also seriously endanger human health.Therefore,finding the appropriate algae coagulation removal process and coagulant injection method is of great significance to ensure the safety of urban drinking water quality.Based on this,this paper studied the prediction of the coagulation removal rate and preoxidation.The main research contents are as follows:(1)The four oxidants of potassium high-speed rate,potassium permanganate,sodium hypochlorite and chlorine dioxide were compared on the promotion effect of algae coagulation and the removal effect of soluble organic matter in the coagulation process.The influence of different oxidants on the algae surface Zeta potential and the flocculation morphology of the algae condensation process were also tested to explore the algae removal mechanism of preoxidation and enhanced coagulation.The results show that the application of oxidant mainly improved the coagulation process and improved the particle size of the morphology of algae and the removal effect of algae.Based on this theory,the algae removal rate prediction based on algae catkins identification is studied.(2)This paper introduces the deep learning algorithm into flocculant image identification.According to the problem of the problem,it proposes a deep learning based algae coagulation removal rate prediction method,using convolutional neural network and image recognition,so as to adjust the dosage.The results show that the prediction accuracy of the improved VGG,Res Net,and Dense Net models is 80.7%,67.3%,and 89.5% for the present dataset,respectively.The Dense Net model has the highest accuracy in transfer learning results for this dataset,while simultaneously adapting the flocculate images produced by aeruginosa microalgae coagulation outside the dataset.In addition,based on the rule of algae coagulation removal and the prediction model of algae coagulation removal rate,we designed a coagulation injection control system equipped with algae coagulation removal rate prediction model,and preliminarily designed the interface.This paper studies the difference of different oxidant coupling,and preliminarily explores the mechanism of pre-oxidation coupling to remove algae,and establishes a set of algae coagulation removal rate prediction model based on deep learning algorithm.Among them,the improved Dense Net network model has high prediction accuracy and excellent generalization,and has high application value and wide application prospect in the automatic coagulation system designed here.