Research On Flotation Agent Dosage Optimization Method And Engineering Realization

Iron ore flotation and dosing is an important link in the process of iron ore flotation,Accurate addition of flotation reagents has a great impact on flotation.In recent years,China’s steel companies lack high-precision flotation dosing equipment and effective dosing strategies in iron ore flotation production,The amount of dosing can often only be added through the experience of the workers.This method of dosing not only increases the labor intensity of workers,but also causes fluctuations in flotation products,iron ore concentrate grades and tailings grades.And it will also cause waste of the dosage,thus affecting the production efficiency of the enterprise.In order to solve the above problems,this study started with a flotation workshop in a sintering plant.Mainly through the analysis of the flotation process of a sinter plant,Establishing a tailing grade classification model based on actual production requirements.Use case-based reasoning to derive dosage.Design and development of flotation dosing automatic system.Apply the system to actual production,The main research contents of this paper are as follows:Firstly,analysis of the entire flotation process,and determined that the flotation effect is a function of the color of the tailings slurry,workers adjust the dosage according to the color of the foam.Using the RGB color characteristics of the slurry,the grade of the flotation tailings can be classified,and the results of the classification can be used to control the dosage of the medicament.This paper uses three methods to establish a classification model for flotation tailings grade.Established RBF neural network tailings grade classification model.Comparison with BP neural network and support vector machine,obtained the superiority of RBF neural network in tailing grade classification.Then,based on the obtained grades and other important factors affecting flotation,a casebased reasoning method was used to establish a model for the amount of flotation reagents.he flotation process,a complex physical and chemical reaction,is hard to be established as a mechanism model.Furthermore,the control of the number of flotation reagents basically relies on the artificial experience due to the fact that the selection of refractory ore containing magnetite and hematite,the frequent fluctuations of ore grade,the lagging technology of infrastructure as well as the comparably-low automation.In this dissertation,the flotation reagent consumption model is built by using the case-based reasoning method,which thinks the operating conditions and process index target value as the case description,regards the number of flotation reagents as the case solution,and through the case retrieval as well as case reuse,the corresponding case solution of the current situation is obtained,namely the number of each flotation regent.Apart from this,the case correction and case storage process set.Finally,designed a software and hardware platform for the flotation automatic dosing system.Completed automatic setting and addition of flotation reagents,changed the history of flotation of a sinter plant by manual addition.The application of the system to the actual industrial site has further verified the results of this research and brought economic benefits to the company.